id string | paper_id string | forum_id null | family_id string | author_component_id string | dataset string | condition string | view string | text string | regions list | target_start int64 | target_end int64 | original_target string | generated_target string | label string | target_label int64 | text_sha256 string | conference string | year int64 | pdf_url string | pdf_sha256 string | clean_prose null | eligible_clean_novel_control null | label_policy string | development_exposed bool | requested_sentence_count int64 | generated_sentence_count int64 | quality_stratum bool | generator string | split string | authors list | title string | metadata_json string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
icml-2016-f0cb38ba3061a21a/g01/sentence/context | icml-2016-f0cb38ba3061a21a | null | icml-2016-f0cb38ba3061a21a/g01 | authors-458cf6806e22b73b | paper_workflow_reconstruction | sentence | context | To identify the best response model for HCPF, we ran SVI with all seven additive EDM distributions in Table 1. We first fit all HCPF models by sampling from the full matrix. We calculated L, L M , L N M and L CN M . In Section 5.3, we compare HCPF and HPF in L. In the second analysis, we only used the non-missing entri... | [
{
"start": 0,
"end": 781,
"label": 0
},
{
"start": 781,
"end": 878,
"label": 1
},
{
"start": 878,
"end": 1669,
"label": 0
}
] | 781 | 878 | We fit HCPF with normal, gamma, and inverse Gaussian as element distributions for all the data sets. | We also fit HCPF with normal, gamma, and inverse Gaussian element distributions on all data sets. | mixed | 1 | 3fd4603d3e928784bd54fcc5c7baa941d1c66e5384767f14f645d9d32343f575 | ICML | 2,016 | https://proceedings.mlr.press/v48/basbug16.pdf | 9e8ac0a9da67099224327462f7efe1407ea859e5e74aa1244a09cad466ef4a88 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Mehmet Basbug",
"Barbara Engelhardt"
] | Hierarchical Compound Poisson Factorization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 2, "candidate_start": 0, "candidate_end": 2}, {"tag": "insert", "original_start": 2, "original_end": 2, "candidate_start": 2, "candidate_end": 7}, {"tag": "equal", "origi... |
icml-2016-f0cb38ba3061a21a/g01/sentence/paragraph | icml-2016-f0cb38ba3061a21a | null | icml-2016-f0cb38ba3061a21a/g01 | authors-458cf6806e22b73b | paper_workflow_reconstruction | sentence | paragraph | In this analysis we quantify how well these models capture both the sparsity and response behavior in ultra sparse matrices. In a movie ratings data set, the question becomes ‘Can we predict if a user would rate a given movie and if she does what rating she would give?’. We report the test log likelihood of all twelve ... | [
{
"start": 0,
"end": 331,
"label": 0
},
{
"start": 331,
"end": 428,
"label": 1
},
{
"start": 428,
"end": 577,
"label": 0
}
] | 331 | 428 | We fit HCPF with normal, gamma, and inverse Gaussian as element distributions for all the data sets. | We also fit HCPF with normal, gamma, and inverse Gaussian element distributions on all data sets. | mixed | 1 | 55132244a7a901c8c86475f5355c470a72a03aa2070af401a38253ce51425aee | ICML | 2,016 | https://proceedings.mlr.press/v48/basbug16.pdf | 9e8ac0a9da67099224327462f7efe1407ea859e5e74aa1244a09cad466ef4a88 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Mehmet Basbug",
"Barbara Engelhardt"
] | Hierarchical Compound Poisson Factorization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 2, "candidate_start": 0, "candidate_end": 2}, {"tag": "insert", "original_start": 2, "original_end": 2, "candidate_start": 2, "candidate_end": 7}, {"tag": "equal", "origi... |
acl-2021-dd3b8f44991d56f2/g01/sentence/context | acl-2021-dd3b8f44991d56f2 | null | acl-2021-dd3b8f44991d56f2/g01 | authors-9456755878a8afa3 | paper_workflow_reconstruction | sentence | context | Enhanced UD A tactic used by many recent releases of UD treebanks is to introduce certain extra edges and non-lexical nodes (Schuster and Manning, 2016; Nivre et al., 2018; Bouma et al., 2020). While some of the theoretical issues still persist in this approach with respect to capturing the symmetric nature of relation... | [
{
"start": 0,
"end": 1568,
"label": 0
},
{
"start": 1568,
"end": 1705,
"label": 1
},
{
"start": 1705,
"end": 1988,
"label": 0
}
] | 1,568 | 1,705 | Conjuncts immediately within the bubble may co-head the bubble, and the bubble itself may establish dependencies with its governor and modifiers. | Conjuncts immediately within the bubble may co-head it, and the bubble itself may establish dependencies with its governor and modifiers. | mixed | 1 | 364d61c20e307d7eb5f70b89ab3268320d262642ac154f20f43afeba15f07ade | ACL | 2,021 | https://aclanthology.org/2021.acl-long.557.pdf | 93d1811950637b6994f203454a1adebd3c02f298984db6260d0bc4d927eae616 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Tianze Shi",
"Lillian Lee"
] | Transition-based Bubble Parsing: Improvements on Coordination Structure Prediction | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 52, "candidate_start": 0, "candidate_end": 52}, {"tag": "insert", "original_start": 52, "original_end": 52, "candidate_start": 52, "candidate_end": 53}, {"tag": "equal", ... |
acl-2021-dd3b8f44991d56f2/g01/sentence/paragraph | acl-2021-dd3b8f44991d56f2 | null | acl-2021-dd3b8f44991d56f2/g01 | authors-9456755878a8afa3 | paper_workflow_reconstruction | sentence | paragraph | An alternative solution to the coordination-independency-trees dilemma is to permit certain restricted phrase-inspired constructs for such structures. Indeed, Tesnière’s (1959) seminal work on dependency grammar does not describe all syntactic relations in terms of dependencies, but rather reserves a primitive relation... | [
{
"start": 0,
"end": 758,
"label": 0
},
{
"start": 758,
"end": 895,
"label": 1
},
{
"start": 895,
"end": 936,
"label": 0
}
] | 758 | 895 | Conjuncts immediately within the bubble may co-head the bubble, and the bubble itself may establish dependencies with its governor and modifiers. | Conjuncts immediately within the bubble may co-head it, and the bubble itself may establish dependencies with its governor and modifiers. | mixed | 1 | 201be6f4874e417ec6946bd9fbf714279365a6ee373b9c1c4000bfcd84b3dce1 | ACL | 2,021 | https://aclanthology.org/2021.acl-long.557.pdf | 93d1811950637b6994f203454a1adebd3c02f298984db6260d0bc4d927eae616 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Tianze Shi",
"Lillian Lee"
] | Transition-based Bubble Parsing: Improvements on Coordination Structure Prediction | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 52, "candidate_start": 0, "candidate_end": 52}, {"tag": "insert", "original_start": 52, "original_end": 52, "candidate_start": 52, "candidate_end": 53}, {"tag": "equal", ... |
2d6e6b9675fb31f6c5250b7ea73fc37d/g01/sentence/context | 2d6e6b9675fb31f6c5250b7ea73fc37d | null | 2d6e6b9675fb31f6c5250b7ea73fc37d/g01 | authors-cf5e199128e09cd7 | paper_workflow_reconstruction | sentence | context | We present two example model specifications in Augur, latent Dirichlet allocation (LDA) [5], and a multivariate linear regression model. The supplementary material shows how to generate samples from the models, and how to use them for prediction. It also contains six more example probabilistic models in Augur: polynomi... | [
{
"start": 0,
"end": 807,
"label": 0
},
{
"start": 807,
"end": 1020,
"label": 1
},
{
"start": 1020,
"end": 2295,
"label": 0
}
] | 807 | 1,020 | The support of the LDA model is composed of four arrays, one each for the distribution of topics per document (theta), the distribution of words per topic (phi), the topics assignments (z), and the words in the corpus (w). | The support of the LDA model consists of four arrays representing the distribution of topics per document (theta), the distribution of words per topic (phi), topic assignments (z), and the words in the corpus (w). | mixed | 1 | 89532871a3861a83010a104150688239d5c07bc28d2d47091b61cae9cdadb0d5 | NeurIPS | 2,014 | https://proceedings.neurips.cc/paper_files/paper/2014/file/2d6e6b9675fb31f6c5250b7ea73fc37d-Paper.pdf | 6b17dab6da106f226140ab743ce870e64b05313e328cb0aa362d916c885b1b4a | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Tristan, Jean-Baptiste",
"Huang, Daniel",
"Tassarotti, Joseph",
"Pocock, Adam",
"Green, Stephen J.",
"Steele, Guy L."
] | Augur: Data-Parallel Probabilistic Modeling | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 29, "candidate_start": 0, "candidate_end": 29}, {"tag": "insert", "original_start": 29, "original_end": 29, "candidate_start": 29, "candidate_end": 33}, {"tag": "equal", ... |
2d6e6b9675fb31f6c5250b7ea73fc37d/g01/sentence/paragraph | 2d6e6b9675fb31f6c5250b7ea73fc37d | null | 2d6e6b9675fb31f6c5250b7ea73fc37d/g01 | authors-cf5e199128e09cd7 | paper_workflow_reconstruction | sentence | paragraph | The LDA model specification is shown in Figure 1a. The probability distribution is a Scala object (object LDA) composed of two declarations. First, we declare the support of the probability distribution as a class named sig. The support of the LDA model consists of four arrays representing the distribution of topics pe... | [
{
"start": 0,
"end": 225,
"label": 0
},
{
"start": 225,
"end": 438,
"label": 1
},
{
"start": 438,
"end": 1434,
"label": 0
}
] | 225 | 438 | The support of the LDA model is composed of four arrays, one each for the distribution of topics per document (theta), the distribution of words per topic (phi), the topics assignments (z), and the words in the corpus (w). | The support of the LDA model consists of four arrays representing the distribution of topics per document (theta), the distribution of words per topic (phi), topic assignments (z), and the words in the corpus (w). | mixed | 1 | 79f695c670ca03815391fa88c2d98e3c6997d7176ee8045fbeaa2ae6151e311e | NeurIPS | 2,014 | https://proceedings.neurips.cc/paper_files/paper/2014/file/2d6e6b9675fb31f6c5250b7ea73fc37d-Paper.pdf | 6b17dab6da106f226140ab743ce870e64b05313e328cb0aa362d916c885b1b4a | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Tristan, Jean-Baptiste",
"Huang, Daniel",
"Tassarotti, Joseph",
"Pocock, Adam",
"Green, Stephen J.",
"Steele, Guy L."
] | Augur: Data-Parallel Probabilistic Modeling | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 29, "candidate_start": 0, "candidate_end": 29}, {"tag": "insert", "original_start": 29, "original_end": 29, "candidate_start": 29, "candidate_end": 33}, {"tag": "equal", ... |
002302d5a1c66195b6981e33e38df11d/g01/sentence/context | 002302d5a1c66195b6981e33e38df11d | null | 002302d5a1c66195b6981e33e38df11d/g01 | authors-65f6f66811a4e363 | paper_workflow_reconstruction | sentence | context | Much of the recent work in obtaining faster learning rates in agnostic learning has taken place in settings where a Bernstein condition holds, including results based on local Rademacher complexities [3, 10]. The Bernstein condition appears to have first been used by Bartlett and Mendelson [4] in their analysis of ERM;... | [
{
"start": 0,
"end": 1411,
"label": 0
},
{
"start": 1411,
"end": 1541,
"label": 1
},
{
"start": 1541,
"end": 2774,
"label": 0
}
] | 1,411 | 1,541 | Secondly, we establish intermediate learning rates that interpolate between the fast and slow rate under a weaker notion of stochastic mixability. | Secondly, we establish intermediate learning rates between the fast and slow rates under a weaker notion of stochastic mixability. | mixed | 1 | a214ae0de06565c3e03cdd0d44874ba5fba69313add0b7b4d173d3929ec4aff4 | NeurIPS | 2,014 | https://proceedings.neurips.cc/paper_files/paper/2014/file/002302d5a1c66195b6981e33e38df11d-Paper.pdf | 74f44e0ef19335d4d5038a54b8ca73470e1aac49bb33b6622f244b7daa721825 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Mehta, Nishant A.",
"Williamson, Robert C."
] | From Stochastic Mixability to Fast Rates | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 51, "candidate_start": 0, "candidate_end": 51}, {"tag": "delete", "original_start": 51, "original_end": 68, "candidate_start": 51, "candidate_end": 51}, {"tag": "equal", ... |
002302d5a1c66195b6981e33e38df11d/g01/sentence/paragraph | 002302d5a1c66195b6981e33e38df11d | null | 002302d5a1c66195b6981e33e38df11d/g01 | authors-65f6f66811a4e363 | paper_workflow_reconstruction | sentence | paragraph | Contributions. The core contribution of this work is to show a new path to the Õ(1/n)-fast rate in statistical learning. We are not aware of previous results that show fast rates from the stochastic mixability assumption. Secondly, we establish intermediate learning rates between the fast and slow rates under a weaker ... | [
{
"start": 0,
"end": 222,
"label": 0
},
{
"start": 222,
"end": 352,
"label": 1
},
{
"start": 352,
"end": 481,
"label": 0
}
] | 222 | 352 | Secondly, we establish intermediate learning rates that interpolate between the fast and slow rate under a weaker notion of stochastic mixability. | Secondly, we establish intermediate learning rates between the fast and slow rates under a weaker notion of stochastic mixability. | mixed | 1 | c9381f2268aa0b265b1e59fe45912d87a7d88ef603da3d33e72381ef9edffc30 | NeurIPS | 2,014 | https://proceedings.neurips.cc/paper_files/paper/2014/file/002302d5a1c66195b6981e33e38df11d-Paper.pdf | 74f44e0ef19335d4d5038a54b8ca73470e1aac49bb33b6622f244b7daa721825 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Mehta, Nishant A.",
"Williamson, Robert C."
] | From Stochastic Mixability to Fast Rates | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 51, "candidate_start": 0, "candidate_end": 51}, {"tag": "delete", "original_start": 51, "original_end": 68, "candidate_start": 51, "candidate_end": 51}, {"tag": "equal", ... |
icml-2014-896c59b46b8804d1/g01/sentence/context | icml-2014-896c59b46b8804d1 | null | icml-2014-896c59b46b8804d1/g01 | authors-9bd633c7d24205aa | paper_workflow_reconstruction | sentence | context | By the exhaustive inference, the performance of PCC is significantly improved in most cases. The good performance highlights the importance of exact and efficient inference rules for PCC. Nevertheless, if the desired evaluation criteria are complicated, it is non-trivial to design exact and efficient inference rules. W... | [
{
"start": 0,
"end": 712,
"label": 0
},
{
"start": 712,
"end": 1126,
"label": 1
},
{
"start": 1126,
"end": 1765,
"label": 0
}
] | 712 | 1,126 | We proposed the condensed filter tree (CFT) algorithm by coupling several tools and ideas: the label powerset approach for reducing to cost-sensitive classification, the tree-based algorithms for cost-sensitive classification, the proper-ordering and K-classifier tricks that utilize the structural property of multi-lab... | We proposed the condensed filter tree (CFT) by coupling several tools and ideas: the label powerset approach for reducing to cost-sensitive classification, tree-based algorithms for cost-sensitive classification, the proper-ordering trick and the K-classifier trick, both of which use a structural property of multi-labe... | mixed | 1 | 41fe04d532cb5fdf3ed0975db46e1ba3b86c7906b66c5ce2ad27ab4a5a7ef1b3 | ICML | 2,014 | https://proceedings.mlr.press/v32/lia14.pdf | eb7c1c48153e56df63454a10bfa579ec3afee1469818ec86a72137bd4acf4c54 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Chun-Liang Li",
"Hsuan-Tien Lin"
] | Condensed Filter Tree for Cost-Sensitive Multi-Label Classification | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 43, "candidate_start": 0, "candidate_end": 43}, {"tag": "delete", "original_start": 43, "original_end": 53, "candidate_start": 43, "candidate_end": 43}, {"tag": "equal", ... |
icml-2014-896c59b46b8804d1/g01/sentence/paragraph | icml-2014-896c59b46b8804d1 | null | icml-2014-896c59b46b8804d1/g01 | authors-9bd633c7d24205aa | paper_workflow_reconstruction | sentence | paragraph | We tackle the general cost-sensitive multi-label classification problem without any specific subroutine for different evaluation criteria, which meets the demands in real-world applications. We proposed the condensed filter tree (CFT) by coupling several tools and ideas: the label powerset approach for reducing to cost... | [
{
"start": 0,
"end": 191,
"label": 0
},
{
"start": 191,
"end": 605,
"label": 1
},
{
"start": 605,
"end": 990,
"label": 0
}
] | 191 | 605 | We proposed the condensed filter tree (CFT) algorithm by coupling several tools and ideas: the label powerset approach for reducing to cost-sensitive classification, the tree-based algorithms for cost-sensitive classification, the proper-ordering and K-classifier tricks that utilize the structural property of multi-lab... | We proposed the condensed filter tree (CFT) by coupling several tools and ideas: the label powerset approach for reducing to cost-sensitive classification, tree-based algorithms for cost-sensitive classification, the proper-ordering trick and the K-classifier trick, both of which use a structural property of multi-labe... | mixed | 1 | e784268b4f477921f16856f0e86abc1bcbabb0a3dcd003655b7cce90b6ec4ba3 | ICML | 2,014 | https://proceedings.mlr.press/v32/lia14.pdf | eb7c1c48153e56df63454a10bfa579ec3afee1469818ec86a72137bd4acf4c54 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Chun-Liang Li",
"Hsuan-Tien Lin"
] | Condensed Filter Tree for Cost-Sensitive Multi-Label Classification | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 43, "candidate_start": 0, "candidate_end": 43}, {"tag": "delete", "original_start": 43, "original_end": 53, "candidate_start": 43, "candidate_end": 43}, {"tag": "equal", ... |
acl-2013-0d06c4bef92179f4/g01/sentence/context | acl-2013-0d06c4bef92179f4 | null | acl-2013-0d06c4bef92179f4/g01 | authors-98f8e0c8b278fcfb | paper_workflow_reconstruction | sentence | context | Let us briefly review the HMM translation model as a starting point. We are given a sequence of English words e = e 1 , . . . , e I . This model produces distributions over French word sequences f = f 1 , . . . , f J and word alignment vectors a = a 1 , . . . , a J , where a j ∈ [0..J] indicates the English word genera... | [
{
"start": 0,
"end": 566,
"label": 0
},
{
"start": 566,
"end": 759,
"label": 1
},
{
"start": 759,
"end": 1326,
"label": 0
}
] | 566 | 759 | Then for each French word position, first the alignment variable (English word index used to generate the current French word) is selected based on only the prior alignment variable. | At each French word position, the alignment variable is first selected, specifying the English word index that generates the current French word, based only on the preceding alignment variable. | mixed | 1 | dac9d74681ad973f72eb11e8046107fac3282795d7db45f31712726b6c0c0f57 | ACL | 2,013 | https://aclanthology.org/P13-2002.pdf | bcb409924ace9ed0ec8da805faa0b057f5d5ff918c2658ded43fe27cc646903b | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Chris Quirk"
] | Exact Maximum Inference for the Fertility Hidden Markov Model | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 8, "candidate_start": 0, "candidate_end": 2}, {"tag": "equal", "original_start": 8, "original_end": 36, "candidate_start": 2, "candidate_end": 30}, {"tag": "delete", "o... |
acl-2013-0d06c4bef92179f4/g01/sentence/paragraph | acl-2013-0d06c4bef92179f4 | null | acl-2013-0d06c4bef92179f4/g01 | authors-98f8e0c8b278fcfb | paper_workflow_reconstruction | sentence | paragraph | The generative story begins by predicting the number of words in the French sentence (hence the number of elements in the alignment vector). At each French word position, the alignment variable is first selected, specifying the English word index that generates the current French word, based only on the preceding align... | [
{
"start": 0,
"end": 141,
"label": 0
},
{
"start": 141,
"end": 334,
"label": 1
},
{
"start": 334,
"end": 508,
"label": 0
}
] | 141 | 334 | Then for each French word position, first the alignment variable (English word index used to generate the current French word) is selected based on only the prior alignment variable. | At each French word position, the alignment variable is first selected, specifying the English word index that generates the current French word, based only on the preceding alignment variable. | mixed | 1 | 189463e295c1eabc418f555ab154c1d4f00c555ef6e9b0c63131beb20701372f | ACL | 2,013 | https://aclanthology.org/P13-2002.pdf | bcb409924ace9ed0ec8da805faa0b057f5d5ff918c2658ded43fe27cc646903b | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Chris Quirk"
] | Exact Maximum Inference for the Fertility Hidden Markov Model | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 8, "candidate_start": 0, "candidate_end": 2}, {"tag": "equal", "original_start": 8, "original_end": 36, "candidate_start": 2, "candidate_end": 30}, {"tag": "delete", "o... |
icml-2020-609bc1a709ab654f/g01/sentence/context | icml-2020-609bc1a709ab654f | null | icml-2020-609bc1a709ab654f/g01 | authors-095c840e5ff853a8 | paper_workflow_reconstruction | sentence | context | Concept of separability to quantify accuracy-fairness tradeoff in the real world: For a group of people in an observed dataset, we quantify the “separability” into positive and negative class labels using Chernoff information, an information-theoretic approximation to the best exponent of the probability of error in bi... | [
{
"start": 0,
"end": 1837,
"label": 0
},
{
"start": 1837,
"end": 1934,
"label": 1
},
{
"start": 1934,
"end": 3024,
"label": 0
}
] | 1,837 | 1,934 | We also formulate an optimization to show how to go about finding such ideal distributions in practice. | We also formulate an optimization to illustrate how to find such ideal distributions in practice. | mixed | 1 | e2cb4af40d1b42d89274a8720ca70aa6166f3cd25fcb286cf3286def39d81dcc | ICML | 2,020 | https://proceedings.mlr.press/v119/dutta20a/dutta20a.pdf | 19fe8ef3e6c6a88eca98e5d12c1003333cc3a467adde8096acf047e3afb7a9fc | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Sanghamitra Dutta",
"Dennis Wei",
"Hazar Yueksel",
"Pin-Yu Chen",
"Sijia Liu",
"Kush Varshney"
] | Is There a Trade-Off Between Fairness and Accuracy? A Perspective Using Mismatched Hypothesis Testing | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 37, "candidate_start": 0, "candidate_end": 37}, {"tag": "insert", "original_start": 37, "original_end": 37, "candidate_start": 37, "candidate_end": 41}, {"tag": "equal", ... |
icml-2020-609bc1a709ab654f/g01/sentence/paragraph | icml-2020-609bc1a709ab654f | null | icml-2020-609bc1a709ab654f/g01 | authors-095c840e5ff853a8 | paper_workflow_reconstruction | sentence | paragraph | Ideal distributions where fairness and accuracy are in accord: Novel to this work, we examine the problem of fair classification through the lens of mismatched hypothesis testing. We show (in Theorem 2) that there exist ideal distributions such that both fairness (in the sense of equal opportunity on both the existing ... | [
{
"start": 0,
"end": 419,
"label": 0
},
{
"start": 419,
"end": 516,
"label": 1
},
{
"start": 516,
"end": 882,
"label": 0
}
] | 419 | 516 | We also formulate an optimization to show how to go about finding such ideal distributions in practice. | We also formulate an optimization to illustrate how to find such ideal distributions in practice. | mixed | 1 | a583d4e5e78d38a578ee95a0a5223c4ed103f9d21450a0b1e8e854d6b6cf739b | ICML | 2,020 | https://proceedings.mlr.press/v119/dutta20a/dutta20a.pdf | 19fe8ef3e6c6a88eca98e5d12c1003333cc3a467adde8096acf047e3afb7a9fc | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Sanghamitra Dutta",
"Dennis Wei",
"Hazar Yueksel",
"Pin-Yu Chen",
"Sijia Liu",
"Kush Varshney"
] | Is There a Trade-Off Between Fairness and Accuracy? A Perspective Using Mismatched Hypothesis Testing | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 37, "candidate_start": 0, "candidate_end": 37}, {"tag": "insert", "original_start": 37, "original_end": 37, "candidate_start": 37, "candidate_end": 41}, {"tag": "equal", ... |
acl-2021-165437127f2486e2/g01/sentence/context | acl-2021-165437127f2486e2 | null | acl-2021-165437127f2486e2/g01 | authors-908ccc76d057130f | paper_workflow_reconstruction | sentence | context | Retrieval-based report generation Retrievalbased approaches are usually hybridized with generation-based ones to improve the readability of generated medical reports. For example, KERP (Li et al., 2019) uses abnormality graphs to retrieve most related sentence templates during the generation. HRGR-Agent(Li et al., 2018... | [
{
"start": 0,
"end": 1166,
"label": 0
},
{
"start": 1166,
"end": 1385,
"label": 1
},
{
"start": 1385,
"end": 2250,
"label": 0
}
] | 1,166 | 1,385 | Finally, MedWriter generates accurate, diverse, and disease-specified medical reports by a hierarchical language decoder that fuses the visual, linguistics and pathological information obtained by VLR and LLR modules. | Finally, MedWriter uses a hierarchical language decoder to fuse visual, linguistics, and pathological information obtained by the VLR and LLR modules, generating accurate, diverse, and disease-specified medical reports. | mixed | 1 | 53b25138fc3bd483b13ff8441ef84df3cbac4e32685a65d3ee58d7fc0f9e8189 | ACL | 2,021 | https://aclanthology.org/2021.acl-long.387.pdf | 0c469e9a4e79e37aa4da213cb8be113d415d8a08b9209a9823d940a70e9713da | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Xingyi Yang",
"Muchao Ye",
"Quanzeng You",
"Fenglong Ma"
] | Writing by Memorizing: Hierarchical Retrieval-based Medical Report Generation | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 19, "candidate_start": 0, "candidate_end": 19}, {"tag": "insert", "original_start": 19, "original_end": 19, "candidate_start": 19, "candidate_end": 151}, {"tag": "equal",... |
acl-2021-165437127f2486e2/g01/sentence/paragraph | acl-2021-165437127f2486e2 | null | acl-2021-165437127f2486e2/g01 | authors-908ccc76d057130f | paper_workflow_reconstruction | sentence | paragraph | As shown in Figure 2, we propose a new framework called MedWriter, which consists of three modules. The Visual-Language Retrieval (VLR) module works on the report level and uses visual features to find the most relevant template reports based on a multi-view image query. The Language-Language Retrieval (LLR) module wor... | [
{
"start": 0,
"end": 489,
"label": 0
},
{
"start": 489,
"end": 708,
"label": 1
},
{
"start": 708,
"end": 875,
"label": 0
}
] | 489 | 708 | Finally, MedWriter generates accurate, diverse, and disease-specified medical reports by a hierarchical language decoder that fuses the visual, linguistics and pathological information obtained by VLR and LLR modules. | Finally, MedWriter uses a hierarchical language decoder to fuse visual, linguistics, and pathological information obtained by the VLR and LLR modules, generating accurate, diverse, and disease-specified medical reports. | mixed | 1 | ffa003bb333648abae137341fbb6f67e2b0e69a01a3fffb0b29906c43ba191d9 | ACL | 2,021 | https://aclanthology.org/2021.acl-long.387.pdf | 0c469e9a4e79e37aa4da213cb8be113d415d8a08b9209a9823d940a70e9713da | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Xingyi Yang",
"Muchao Ye",
"Quanzeng You",
"Fenglong Ma"
] | Writing by Memorizing: Hierarchical Retrieval-based Medical Report Generation | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 19, "candidate_start": 0, "candidate_end": 19}, {"tag": "insert", "original_start": 19, "original_end": 19, "candidate_start": 19, "candidate_end": 151}, {"tag": "equal",... |
icml-2021-1428317ece72be19/g01/sentence/context | icml-2021-1428317ece72be19 | null | icml-2021-1428317ece72be19/g01 | authors-53ae0961205bafb6 | paper_workflow_reconstruction | sentence | context | In this work, we prove that, up to universal multiplicative constants and additive log(K) terms in the exponential, the optimal error probability is exp(−T min k ∆ 2 k ), which highlights the somewhat surprising fact that this structured concave TBP problem is also akin to a one armed TBP- see Subsection 3.3. We provid... | [
{
"start": 0,
"end": 586,
"label": 0
},
{
"start": 586,
"end": 694,
"label": 1
},
{
"start": 694,
"end": 1282,
"label": 0
}
] | 586 | 694 | We also describe the performance criterion - probability of error, we will be primarily using for the duration of the paper. | We will also primarily refer to this performance criterion as the probability of error throughout the paper. | mixed | 1 | 1337ac1af7827373746cb8fc2dcfcc826b5157d3a615ab383d3898d8bc5f3848 | ICML | 2,021 | https://proceedings.mlr.press/v139/cheshire21a/cheshire21a.pdf | 98da4617a9377317583f41a3440074364d5386e9cbc2a6c23ad7d88ce33f0dd3 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"James Cheshire",
"Pierre Menard",
"Alexandra Carpentier"
] | Problem Dependent View on Structured Thresholding Bandit Problems | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 2, "candidate_start": 0, "candidate_end": 2}, {"tag": "insert", "original_start": 2, "original_end": 2, "candidate_start": 2, "candidate_end": 7}, {"tag": "equal", "origi... |
icml-2021-1428317ece72be19/g01/sentence/paragraph | icml-2021-1428317ece72be19 | null | icml-2021-1428317ece72be19/g01 | authors-53ae0961205bafb6 | paper_workflow_reconstruction | sentence | paragraph | Organisation of the paper This paper is structured as follows. In Section 2 we formally introduce the TBP setting along with the monotone and concave shape constraints. We will also primarily refer to this performance criterion as the probability of error throughout the paper. Following this, upper and lower bounds on ... | [
{
"start": 0,
"end": 169,
"label": 0
},
{
"start": 169,
"end": 277,
"label": 1
},
{
"start": 277,
"end": 700,
"label": 0
}
] | 169 | 277 | We also describe the performance criterion - probability of error, we will be primarily using for the duration of the paper. | We will also primarily refer to this performance criterion as the probability of error throughout the paper. | mixed | 1 | 6280ddfbedd7aefe31b0adb8c70a44abc102dd95a2b8d22143ece63cba5d0fb4 | ICML | 2,021 | https://proceedings.mlr.press/v139/cheshire21a/cheshire21a.pdf | 98da4617a9377317583f41a3440074364d5386e9cbc2a6c23ad7d88ce33f0dd3 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"James Cheshire",
"Pierre Menard",
"Alexandra Carpentier"
] | Problem Dependent View on Structured Thresholding Bandit Problems | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 2, "candidate_start": 0, "candidate_end": 2}, {"tag": "insert", "original_start": 2, "original_end": 2, "candidate_start": 2, "candidate_end": 7}, {"tag": "equal", "origi... |
acl-2017-30477ba5253f03ff/g01/sentence/context | acl-2017-30477ba5253f03ff | null | acl-2017-30477ba5253f03ff/g01 | authors-e69898b223791ed6 | paper_workflow_reconstruction | sentence | context | The SICK corpus (Bentivogli et al., 2014) consists of 10K pairs of English sentences containing multiple lexical, syntactic, and semantic phenomena. It builds on two external data sources – the 8K ImageFlickr dataset (Rashtchian et al., 2010) and SemEval-2012 Semantic Textual Similarity dataset (Agirre et al., 2012). E... | [
{
"start": 0,
"end": 1432,
"label": 0
},
{
"start": 1432,
"end": 1537,
"label": 1
},
{
"start": 1537,
"end": 2714,
"label": 0
}
] | 1,432 | 1,537 | In order to verify whether an NLP algorithm is adequate, it is not enough to evaluate it solely for English. | Therefore, evaluating an NLP algorithm only for English is insufficient to verify whether it is adequate. | mixed | 1 | 4d8957079a8cf34d24159c3ae3c8a5d11f8c956969ec86747f4f40183444a14b | ACL | 2,017 | https://aclanthology.org/P17-1073.pdf | 9a37a39c135bba85575e0810084b028dd3fac7229123b50ee4f69cdbdca52482 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Alina Wróblewska",
"Katarzyna Krasnowska-Kieraś"
] | Polish evaluation dataset for compositional distributional semantics models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 1, "candidate_start": 0, "candidate_end": 23}, {"tag": "equal", "original_start": 1, "original_end": 3, "candidate_start": 23, "candidate_end": 25}, {"tag": "insert", "... |
acl-2017-30477ba5253f03ff/g01/sentence/paragraph | acl-2017-30477ba5253f03ff | null | acl-2017-30477ba5253f03ff/g01 | authors-e69898b223791ed6 | paper_workflow_reconstruction | sentence | paragraph | Studying approaches to various natural language processing (henceforth NLP) problems, we have observed that the availability of language resources (e.g. training or testing data) stimulates the development of NLP tools and the estimation of NLP models. English is undoubtedly the most prominent in this regard and Englis... | [
{
"start": 0,
"end": 493,
"label": 0
},
{
"start": 493,
"end": 598,
"label": 1
},
{
"start": 598,
"end": 1005,
"label": 0
}
] | 493 | 598 | In order to verify whether an NLP algorithm is adequate, it is not enough to evaluate it solely for English. | Therefore, evaluating an NLP algorithm only for English is insufficient to verify whether it is adequate. | mixed | 1 | 762ee87bdd45ac6816d35c0b8350ad61733a0c67a22256e52a2187fecbb27a20 | ACL | 2,017 | https://aclanthology.org/P17-1073.pdf | 9a37a39c135bba85575e0810084b028dd3fac7229123b50ee4f69cdbdca52482 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Alina Wróblewska",
"Katarzyna Krasnowska-Kieraś"
] | Polish evaluation dataset for compositional distributional semantics models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 1, "candidate_start": 0, "candidate_end": 23}, {"tag": "equal", "original_start": 1, "original_end": 3, "candidate_start": 23, "candidate_end": 25}, {"tag": "insert", "... |
e02af5824e1eb6ad58d6bc03ac9e827f/g01/sentence/context | e02af5824e1eb6ad58d6bc03ac9e827f | null | e02af5824e1eb6ad58d6bc03ac9e827f/g01 | authors-85ebffaaa937ef5e | paper_workflow_reconstruction | sentence | context | Computational neuroscience—the study of how neural circuits transform sensory inputs into behavioral outputs—is intimately coupled with computational ethology—the quantitative analysis of behavior [1, 2]. In order to understand the computations of the nervous system, we must first have a rigorous description of the beh... | [
{
"start": 0,
"end": 942,
"label": 0
},
{
"start": 942,
"end": 1097,
"label": 1
},
{
"start": 1097,
"end": 1708,
"label": 0
}
] | 942 | 1,097 | They swim with brief tail flicks, or bouts, that propel them forward, reorient them, and enable them to pursue and capture prey [5, 6]. | Larval zebrafish swim through brief tail flicks, termed bouts, which propel them forward, reorient them, and enable them to pursue and capture prey [5, 6]. | mixed | 1 | bded9589e9cd480e0505b6de2b24df7a649c59b4bad7711ae4a157fb4f1aed44 | NeurIPS | 2,018 | https://proceedings.neurips.cc/paper_files/paper/2018/file/e02af5824e1eb6ad58d6bc03ac9e827f-Paper.pdf | 8a38c9e09b254712fcbae3776559c756bd074bc48c2ae9d2c7b9787faa83272b | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Sharma, Anuj",
"Johnson, Robert",
"Engert, Florian",
"Linderman, Scott"
] | Point process latent variable models of larval zebrafish behavior | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 1, "candidate_start": 0, "candidate_end": 15}, {"tag": "equal", "original_start": 1, "original_end": 2, "candidate_start": 15, "candidate_end": 16}, {"tag": "delete", "... |
e02af5824e1eb6ad58d6bc03ac9e827f/g01/sentence/paragraph | e02af5824e1eb6ad58d6bc03ac9e827f | null | e02af5824e1eb6ad58d6bc03ac9e827f/g01 | authors-85ebffaaa937ef5e | paper_workflow_reconstruction | sentence | paragraph | For many organisms, overt behaviors manifest as a sequence of discrete and nearly-instantaneous events unfolding over time, often with some associated measurements, or marks. Multiple times a second, our eyes saccade in a quick, jerking motion to fixate on a new point in our field of view [3]. Some electric fish emit p... | [
{
"start": 0,
"end": 466,
"label": 0
},
{
"start": 466,
"end": 621,
"label": 1
},
{
"start": 621,
"end": 889,
"label": 0
}
] | 466 | 621 | They swim with brief tail flicks, or bouts, that propel them forward, reorient them, and enable them to pursue and capture prey [5, 6]. | Larval zebrafish swim through brief tail flicks, termed bouts, which propel them forward, reorient them, and enable them to pursue and capture prey [5, 6]. | mixed | 1 | b5f3b29ed471568a6f50987a8a33842d2ab7c9e08e5801aa35e14ab097f257c3 | NeurIPS | 2,018 | https://proceedings.neurips.cc/paper_files/paper/2018/file/e02af5824e1eb6ad58d6bc03ac9e827f-Paper.pdf | 8a38c9e09b254712fcbae3776559c756bd074bc48c2ae9d2c7b9787faa83272b | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Sharma, Anuj",
"Johnson, Robert",
"Engert, Florian",
"Linderman, Scott"
] | Point process latent variable models of larval zebrafish behavior | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 1, "candidate_start": 0, "candidate_end": 15}, {"tag": "equal", "original_start": 1, "original_end": 2, "candidate_start": 15, "candidate_end": 16}, {"tag": "delete", "... |
0f3d014eead934bbdbacb62a01dc4831/g01/sentence/context | 0f3d014eead934bbdbacb62a01dc4831 | null | 0f3d014eead934bbdbacb62a01dc4831/g01 | authors-3c772f8af046365c | paper_workflow_reconstruction | sentence | context | In a somewhat tangential direction, random neural networks have also been investigated through their relationship to kernel methods. The correspondence between infinite-dimensional neural networks and Gaussian processes was first noted by Neal (1994a,b). In the finite-dimensional setting, the approximate correspondence... | [
{
"start": 0,
"end": 825,
"label": 0
},
{
"start": 825,
"end": 934,
"label": 1
},
{
"start": 934,
"end": 1536,
"label": 0
}
] | 825 | 934 | Schoenholz et al. (2016) studied how information propagates through random networks, and how that affects learning. | Schoenholz et al. (2016) studied information propagation through random networks and how it affects learning. | mixed | 1 | 13d43b48524c8704bbbdf9cb2376ab3dec3f269ec182b4d8fddbc550bdd579b1 | NeurIPS | 2,017 | https://proceedings.neurips.cc/paper_files/paper/2017/file/0f3d014eead934bbdbacb62a01dc4831-Paper.pdf | b0d5e50d8f346d83e3bebe2deeba8edd01a962e33dbb1e8b2c40be849f845e09 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Pennington, Jeffrey",
"Worah, Pratik"
] | Nonlinear random matrix theory for deep learning | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 33, "candidate_start": 0, "candidate_end": 33}, {"tag": "delete", "original_start": 33, "original_end": 37, "candidate_start": 33, "candidate_end": 33}, {"tag": "equal", ... |
0f3d014eead934bbdbacb62a01dc4831/g01/sentence/paragraph | 0f3d014eead934bbdbacb62a01dc4831 | null | 0f3d014eead934bbdbacb62a01dc4831/g01 | authors-3c772f8af046365c | paper_workflow_reconstruction | sentence | paragraph | In the last several years, random neural networks have been studied from many other perspectives. Saxe et al. (2014) examined the effect of random initialization on the dynamics of learning in deep linear networks. Schoenholz et al. (2016) studied information propagation through random networks and how it affects learn... | [
{
"start": 0,
"end": 215,
"label": 0
},
{
"start": 215,
"end": 324,
"label": 1
},
{
"start": 324,
"end": 461,
"label": 0
}
] | 215 | 324 | Schoenholz et al. (2016) studied how information propagates through random networks, and how that affects learning. | Schoenholz et al. (2016) studied information propagation through random networks and how it affects learning. | mixed | 1 | e131e06797effc303b4856aa99923de17b78ae1e68f4c54f4e8125b0fe7ed54a | NeurIPS | 2,017 | https://proceedings.neurips.cc/paper_files/paper/2017/file/0f3d014eead934bbdbacb62a01dc4831-Paper.pdf | b0d5e50d8f346d83e3bebe2deeba8edd01a962e33dbb1e8b2c40be849f845e09 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Pennington, Jeffrey",
"Worah, Pratik"
] | Nonlinear random matrix theory for deep learning | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 33, "candidate_start": 0, "candidate_end": 33}, {"tag": "delete", "original_start": 33, "original_end": 37, "candidate_start": 33, "candidate_end": 33}, {"tag": "equal", ... |
c52f1bd66cc19d05628bd8bf27af3ad6/g01/sentence/context | c52f1bd66cc19d05628bd8bf27af3ad6 | null | c52f1bd66cc19d05628bd8bf27af3ad6/g01 | authors-f627b96b1a6ec8c3 | paper_workflow_reconstruction | sentence | context | Area Under the ROC Curve (AUC) [8] is a widely used metric for measuring classification performance. Unlike misclassification error that reflects a classifier’s ability to classify a single randomly chosen example, AUC concerns the overall performance of a functional family of classifiers and quantifies their ability o... | [
{
"start": 0,
"end": 696,
"label": 0
},
{
"start": 696,
"end": 870,
"label": 1
},
{
"start": 870,
"end": 3053,
"label": 0
}
] | 696 | 870 | However, most studies of online learning focus on the misclassification error or its surrogate loss, in which the objective function depends on a sum of losses over individual examples. | However, most studies of online learning focus on misclassification error or its surrogate loss, whose objective function depends on a sum of losses over individual examples. | mixed | 1 | b5b3237ad3c20ccd321034499f4dc4ba4b1dba386b3276c1024219c1865af270 | NeurIPS | 2,016 | https://proceedings.neurips.cc/paper_files/paper/2016/file/c52f1bd66cc19d05628bd8bf27af3ad6-Paper.pdf | d68925348cace070f10ad352a9097a97c4656d321cb38bd8f18f8141c7f79c35 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Ying, Yiming",
"Wen, Longyin",
"Lyu, Siwei"
] | Stochastic Online AUC Maximization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 50, "candidate_start": 0, "candidate_end": 50}, {"tag": "delete", "original_start": 50, "original_end": 54, "candidate_start": 50, "candidate_end": 50}, {"tag": "equal", ... |
c52f1bd66cc19d05628bd8bf27af3ad6/g01/sentence/paragraph | c52f1bd66cc19d05628bd8bf27af3ad6 | null | c52f1bd66cc19d05628bd8bf27af3ad6/g01 | authors-f627b96b1a6ec8c3 | paper_workflow_reconstruction | sentence | paragraph | On the other hand, online learning algorithms [1, 2, 3, 16, 19, 22], have been proven to be very efficient to deal with large-scale datasets. However, most studies of online learning focus on misclassification error or its surrogate loss, whose objective function depends on a sum of losses over individual examples. It ... | [
{
"start": 0,
"end": 142,
"label": 0
},
{
"start": 142,
"end": 316,
"label": 1
},
{
"start": 316,
"end": 796,
"label": 0
}
] | 142 | 316 | However, most studies of online learning focus on the misclassification error or its surrogate loss, in which the objective function depends on a sum of losses over individual examples. | However, most studies of online learning focus on misclassification error or its surrogate loss, whose objective function depends on a sum of losses over individual examples. | mixed | 1 | 73f3ce832dd9ce8a40a3074e80c494f70169e742ab2458ac6fc965c683da1f32 | NeurIPS | 2,016 | https://proceedings.neurips.cc/paper_files/paper/2016/file/c52f1bd66cc19d05628bd8bf27af3ad6-Paper.pdf | d68925348cace070f10ad352a9097a97c4656d321cb38bd8f18f8141c7f79c35 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Ying, Yiming",
"Wen, Longyin",
"Lyu, Siwei"
] | Stochastic Online AUC Maximization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 50, "candidate_start": 0, "candidate_end": 50}, {"tag": "delete", "original_start": 50, "original_end": 54, "candidate_start": 50, "candidate_end": 50}, {"tag": "equal", ... |
icml-2018-5d66e9074c5a7df7/g01/sentence/context | icml-2018-5d66e9074c5a7df7 | null | icml-2018-5d66e9074c5a7df7/g01 | authors-f66792c69f10dd51 | paper_workflow_reconstruction | sentence | context | Having reviewed basic properties of Kreı̆n spaces, we are now ready to introduce the notion of a reproducing kernel Kreı̆n space. For that, let X be an instance space and denote with R X the set of functions from X to R.
In contrast to reproducing kernel Hilbert spaces, there is no bijection between reproducing kernel... | [
{
"start": 0,
"end": 371,
"label": 0
},
{
"start": 371,
"end": 521,
"label": 1
},
{
"start": 521,
"end": 1940,
"label": 0
}
] | 371 | 521 | Moreover, it is important to note that not every symmetric kernel function admits a representation as a difference between two positive definite kernels. | Moreover, it is important to note that not every symmetric kernel function admits a representation as the difference of two positive definite kernels. | mixed | 1 | 88e6fc89456cb5e88873a04b91a6c065a4364e39b6167f15ec723d01a8cd12bb | ICML | 2,018 | https://proceedings.mlr.press/v80/oglic18a/oglic18a.pdf | 99ed7cbf8cb18b1534e2379934947d8f44854321a9a03dc18417645495ac99c5 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Dino Oglic",
"Thomas Gaertner"
] | Learning in Reproducing Kernel Kreı̆n Spaces | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 102, "candidate_start": 0, "candidate_end": 102}, {"tag": "replace", "original_start": 102, "original_end": 103, "candidate_start": 102, "candidate_end": 105}, {"tag": "e... |
icml-2018-5d66e9074c5a7df7/g01/sentence/paragraph | icml-2018-5d66e9074c5a7df7 | null | icml-2018-5d66e9074c5a7df7/g01 | authors-f66792c69f10dd51 | paper_workflow_reconstruction | sentence | paragraph | In contrast to reproducing kernel Hilbert spaces, there is no bijection between reproducing kernel Kreı̆n spaces and indefinite reproducing kernels. Moreover, it is important to note that not every symmetric kernel function admits a representation as the difference of two positive definite kernels. A symmetric function... | [
{
"start": 0,
"end": 149,
"label": 0
},
{
"start": 149,
"end": 299,
"label": 1
},
{
"start": 299,
"end": 554,
"label": 0
}
] | 149 | 299 | Moreover, it is important to note that not every symmetric kernel function admits a representation as a difference between two positive definite kernels. | Moreover, it is important to note that not every symmetric kernel function admits a representation as the difference of two positive definite kernels. | mixed | 1 | 93668311709cd47c4e8d89dcac0b8c15033d3f52bc5cfe60fd427726d992e184 | ICML | 2,018 | https://proceedings.mlr.press/v80/oglic18a/oglic18a.pdf | 99ed7cbf8cb18b1534e2379934947d8f44854321a9a03dc18417645495ac99c5 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Dino Oglic",
"Thomas Gaertner"
] | Learning in Reproducing Kernel Kreı̆n Spaces | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 102, "candidate_start": 0, "candidate_end": 102}, {"tag": "replace", "original_start": 102, "original_end": 103, "candidate_start": 102, "candidate_end": 105}, {"tag": "e... |
icml-2014-b8f415cd1dc2c467/g01/sentence/context | icml-2014-b8f415cd1dc2c467 | null | icml-2014-b8f415cd1dc2c467/g01 | authors-1440ed152f3cd673 | paper_workflow_reconstruction | sentence | context | I NCORPORATING USER FEATURES . E DGE E XPLAIN easily generalizes to broader settings with multiple user features, such as group memberships, topics of interest, keywords, or pages liked by the user. As an example, consider group memberships of users. Intuitively, if most members of a group come from the same college, t... | [
{
"start": 0,
"end": 1283,
"label": 0
},
{
"start": 1283,
"end": 1407,
"label": 1
},
{
"start": 1407,
"end": 1710,
"label": 0
}
] | 1,283 | 1,407 | All these situations can be easily handled by modifying Eq. 4 to include an edge-specific and label typespecific weight. | These situations can be easily handled by modifying Eq. 4 to include a weight that is edge-specific and label type-specific. | mixed | 1 | fc705adef207cccae42532596b02533db5bb09e9216076f5b69ee4a0802f79a0 | ICML | 2,014 | https://proceedings.mlr.press/v32/chakrabarti14.pdf | 427e9367091fade1eaa5f6d37b3c0dc8844c40a11499691ed6db78e9f41a62ad | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Deepayan Chakrabarti",
"Stanislav Funiak",
"Jonathan Chang",
"Sofus Macskassy"
] | Joint Inference of Multiple Label Types in Large Networks | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 5, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 5, "original_end": 74, "candidate_start": 1, "candidate_end": 70}, {"tag": "replace", "... |
icml-2014-b8f415cd1dc2c467/g01/sentence/paragraph | icml-2014-b8f415cd1dc2c467 | null | icml-2014-b8f415cd1dc2c467/g01 | authors-1440ed152f3cd673 | paper_workflow_reconstruction | sentence | paragraph | I NCORPORATING EDGE FEATURES . There are several situations where edge-specific features could be useful. First, we may want to give more importance to certain kinds of edges, such as the group-membership edges mentioned above. Second, some features could be important for one label type but not another: e.g., the age d... | [
{
"start": 0,
"end": 406,
"label": 0
},
{
"start": 406,
"end": 530,
"label": 1
},
{
"start": 530,
"end": 599,
"label": 0
}
] | 406 | 530 | All these situations can be easily handled by modifying Eq. 4 to include an edge-specific and label typespecific weight. | These situations can be easily handled by modifying Eq. 4 to include a weight that is edge-specific and label type-specific. | mixed | 1 | f27254109c20a347c81710b6fb61431ead0f19652f04cfed13ae4e3ccc434ba9 | ICML | 2,014 | https://proceedings.mlr.press/v32/chakrabarti14.pdf | 427e9367091fade1eaa5f6d37b3c0dc8844c40a11499691ed6db78e9f41a62ad | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Deepayan Chakrabarti",
"Stanislav Funiak",
"Jonathan Chang",
"Sofus Macskassy"
] | Joint Inference of Multiple Label Types in Large Networks | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 5, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 5, "original_end": 74, "candidate_start": 1, "candidate_end": 70}, {"tag": "replace", "... |
acl-2013-615a2fa7f1d6bb18/g01/sentence/context | acl-2013-615a2fa7f1d6bb18 | null | acl-2013-615a2fa7f1d6bb18/g01 | authors-82a0f57dde1233cb | paper_workflow_reconstruction | sentence | context | The final evaluation of the text normalization system is an extrinsic evaluation where we evaluate the effect of the text normalization task on a social media text translating from English to Spanish using a large scale translation system trained on general domain data. The system is trained on English-Spanish parallel... | [
{
"start": 0,
"end": 1194,
"label": 0
},
{
"start": 1194,
"end": 1321,
"label": 1
},
{
"start": 1321,
"end": 1751,
"label": 0
}
] | 1,194 | 1,321 | We compared the system with conventional correction approaches and with recent previous work; and we showed that it highly outperforms other systems. | We compared our system with conventional correction approaches and recent prior work, and it highly outperformed other systems. | mixed | 1 | 97c1158dd0ede987aef9d2923851048fb0d252462c8c01dbe04b8171bd53d9ad | ACL | 2,013 | https://aclanthology.org/P13-1155.pdf | aec8907cbba86804b0aa8638b76c02af403cbdd3e4ebd093bd17b892210296f7 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Hany Hassan Awadalla",
"Arul Menezes"
] | Social Text Normalization using Contextual Graph Random Walks | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 12, "candidate_start": 0, "candidate_end": 12}, {"tag": "replace", "original_start": 12, "original_end": 15, "candidate_start": 12, "candidate_end": 15}, {"tag": "equal",... |
acl-2013-615a2fa7f1d6bb18/g01/sentence/paragraph | acl-2013-615a2fa7f1d6bb18 | null | acl-2013-615a2fa7f1d6bb18/g01 | authors-82a0f57dde1233cb | paper_workflow_reconstruction | sentence | paragraph | We introduced a social media text normalization system that can be deployed as a preprocessor for MT and various NLP applications to handle social media text. The proposed approach is very scalable, adaptive to any domain and language. We show that the proposed unsupervised approach provides a normalization system with... | [
{
"start": 0,
"end": 366,
"label": 0
},
{
"start": 366,
"end": 493,
"label": 1
},
{
"start": 493,
"end": 629,
"label": 0
}
] | 366 | 493 | We compared the system with conventional correction approaches and with recent previous work; and we showed that it highly outperforms other systems. | We compared our system with conventional correction approaches and recent prior work, and it highly outperformed other systems. | mixed | 1 | 89ce852790502de008701444bb53cd24189e4d6510136205536429de6eea28bc | ACL | 2,013 | https://aclanthology.org/P13-1155.pdf | aec8907cbba86804b0aa8638b76c02af403cbdd3e4ebd093bd17b892210296f7 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Hany Hassan Awadalla",
"Arul Menezes"
] | Social Text Normalization using Contextual Graph Random Walks | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 12, "candidate_start": 0, "candidate_end": 12}, {"tag": "replace", "original_start": 12, "original_end": 15, "candidate_start": 12, "candidate_end": 15}, {"tag": "equal",... |
acl-2019-637513d8c85fbdc6/g01/sentence/context | acl-2019-637513d8c85fbdc6 | null | acl-2019-637513d8c85fbdc6/g01 | authors-3992bc307d031e2a | paper_workflow_reconstruction | sentence | context | F1 (0.7980 vs 0.7618) and Macro-F1 (0.5166 vs 0.4334) scores, respectively. Using binary classifiers with a very basic architecture beats DGCNN easily.
Addition of class weights during model training (GRU-Att) further improves the binary models. We optimize the class weights based on the F1- score on the validation da... | [
{
"start": 0,
"end": 324,
"label": 0
},
{
"start": 324,
"end": 409,
"label": 1
},
{
"start": 409,
"end": 1441,
"label": 0
}
] | 324 | 409 | As can be seen from the table, Macro-F1 improves by close to 10% after incorporating class weights. | As shown in the table, incorporating class weights improves Macro-F1 by close to 10%. | mixed | 1 | b6a75dba5638a15d3c94b554885b3a63678ebbb8c3bd2eaab9e4287077d75038 | ACL | 2,019 | https://aclanthology.org/P19-1633.pdf | bef169f26013084f7065d90e18aa656845e14d5ce05ddd533e5d1b38955ff188 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Siddhartha Banerjee",
"Cem Akkaya",
"Francisco Perez-Sorrosal",
"Kostas Tsioutsiouliklis"
] | Hierarchical Transfer Learning for Multi-label Text Classification | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 3}, {"tag": "replace", "original_start": 3, "original_end": 5, "candidate_start": 3, "candidate_end": 7}, {"tag": "equal", "orig... |
acl-2019-637513d8c85fbdc6/g01/sentence/paragraph | acl-2019-637513d8c85fbdc6 | null | acl-2019-637513d8c85fbdc6/g01 | authors-3992bc307d031e2a | paper_workflow_reconstruction | sentence | paragraph | Addition of class weights during model training (GRU-Att) further improves the binary models. We optimize the class weights based on the F1- score on the validation data. As shown in the table, incorporating class weights improves Macro-F1 by close to 10%. The Micro-F1 remains unchanged, though. Therefore, the biggest ... | [
{
"start": 0,
"end": 171,
"label": 0
},
{
"start": 171,
"end": 256,
"label": 1
},
{
"start": 256,
"end": 435,
"label": 0
}
] | 171 | 256 | As can be seen from the table, Macro-F1 improves by close to 10% after incorporating class weights. | As shown in the table, incorporating class weights improves Macro-F1 by close to 10%. | mixed | 1 | 514c4e40737e2dd0d7d63aceff69b2685ece72136348179ee832be252dd74bed | ACL | 2,019 | https://aclanthology.org/P19-1633.pdf | bef169f26013084f7065d90e18aa656845e14d5ce05ddd533e5d1b38955ff188 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Siddhartha Banerjee",
"Cem Akkaya",
"Francisco Perez-Sorrosal",
"Kostas Tsioutsiouliklis"
] | Hierarchical Transfer Learning for Multi-label Text Classification | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 3}, {"tag": "replace", "original_start": 3, "original_end": 5, "candidate_start": 3, "candidate_end": 7}, {"tag": "equal", "orig... |
icml-2020-33fd49a1c3995576/g01/sentence/context | icml-2020-33fd49a1c3995576 | null | icml-2020-33fd49a1c3995576/g01 | authors-c31461b59e6b6093 | paper_workflow_reconstruction | sentence | context | As an encoding mechanism, CPPNs offer desirable properties for open-endedness: (1) They are typically initialized with simple topologies (e.g. no hidden nodes), and are trained with NEAT (Stanley & Miikkulainen, 2002), a neuroevolution (Stanley et al., 2019) algorithm that learns both the topology and the weights of CP... | [
{
"start": 0,
"end": 1641,
"label": 0
},
{
"start": 1641,
"end": 1865,
"label": 1
},
{
"start": 1865,
"end": 3314,
"label": 0
}
] | 1,641 | 1,865 | The new idea here is that measuring progress can be based on the idea that if an existing set of agents are able to solve all of the new challenges generated by a system in the future, then the system has not generated any meaningfully new challenges. | A new idea here is to base measuring progress on agent capability: if the existing set of agents can solve every new challenge the system generates in the future, then the system has generated no meaningfully new challenges. | mixed | 1 | 3dc53c830fe3f09b0f0d990af589fc9066a26ffbf7e5596018ee703471acfbc7 | ICML | 2,020 | https://proceedings.mlr.press/v119/wang20l/wang20l.pdf | 6a01e35d2d0a6ee93015824d3fd2789997057fd604f9389089d5fe04d28c10ab | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Rui Wang",
"Joel Lehman",
"Aditya Rawal",
"Jiale Zhi",
"Yulun Li",
"Jeffrey Clune",
"Kenneth Stanley"
] | Enhanced POET: Open-ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutions | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 3, "original_end": 22, "candidate_start": 1, "candidate_end": 20}, {"tag": "replace", "... |
icml-2020-33fd49a1c3995576/g01/sentence/paragraph | icml-2020-33fd49a1c3995576 | null | icml-2020-33fd49a1c3995576/g01 | authors-c31461b59e6b6093 | paper_workflow_reconstruction | sentence | paragraph | Measuring progress in open-ended systems has long presented a challenge to pursuing open-endedness: As there is no a priori expected outcome against which progress can be measured, how can we tell whether a system continues to generate interesting new things? A new idea here is to base measuring progress on agent capab... | [
{
"start": 0,
"end": 260,
"label": 0
},
{
"start": 260,
"end": 484,
"label": 1
},
{
"start": 484,
"end": 1423,
"label": 0
}
] | 260 | 484 | The new idea here is that measuring progress can be based on the idea that if an existing set of agents are able to solve all of the new challenges generated by a system in the future, then the system has not generated any meaningfully new challenges. | A new idea here is to base measuring progress on agent capability: if the existing set of agents can solve every new challenge the system generates in the future, then the system has generated no meaningfully new challenges. | mixed | 1 | 577e4a615b4500fbb9ce3073ee3fcb8786c8c9c605ebc3597710cad000fe70ca | ICML | 2,020 | https://proceedings.mlr.press/v119/wang20l/wang20l.pdf | 6a01e35d2d0a6ee93015824d3fd2789997057fd604f9389089d5fe04d28c10ab | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Rui Wang",
"Joel Lehman",
"Aditya Rawal",
"Jiale Zhi",
"Yulun Li",
"Jeffrey Clune",
"Kenneth Stanley"
] | Enhanced POET: Open-ended Reinforcement Learning through Unbounded Invention of Learning Challenges and their Solutions | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 3, "original_end": 22, "candidate_start": 1, "candidate_end": 20}, {"tag": "replace", "... |
e96c7de8f6390b1e6c71556e4e0a4959/g01/sentence/context | e96c7de8f6390b1e6c71556e4e0a4959 | null | e96c7de8f6390b1e6c71556e4e0a4959/g01 | authors-7ba9f4d0c4cfcc3f | paper_workflow_reconstruction | sentence | context | The aforementioned CSs and CIs have a strong relationship with corresponding hypothesis tests. In the following section, we discuss how one can use the techniques developed here to sequentially test hypotheses about finite sets of nonrandom numbers.
In classical hypothesis testing, one has access to i.i.d. data from s... | [
{
"start": 0,
"end": 617,
"label": 0
},
{
"start": 617,
"end": 678,
"label": 1
},
{
"start": 678,
"end": 1146,
"label": 0
}
] | 617 | 678 | Recalling the setup of Section 1.1, this is the nonstandard setting we find ourselves in. | In the setup of Section 1.1, we are in a nonstandard setting. | mixed | 1 | 3605777fc8414915040fea2a7c3c2f8b4af7cb15d0bc0e69138b2bf033624379 | NeurIPS | 2,020 | https://proceedings.neurips.cc/paper_files/paper/2020/file/e96c7de8f6390b1e6c71556e4e0a4959-Paper.pdf | 3a3c8cdb16186b6211faf5d287af949cfecab12ef3f57baaad9a634fb4fd5952 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Waudby-Smith, Ian",
"Ramdas, Aaditya"
] | Confidence sequences for sampling without replacement | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 7, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 7, "original_end": 8, "candidate_start": 1, "candidate_end": 2}, {"tag": "delete", "ori... |
e96c7de8f6390b1e6c71556e4e0a4959/g01/sentence/paragraph | e96c7de8f6390b1e6c71556e4e0a4959 | null | e96c7de8f6390b1e6c71556e4e0a4959/g01 | authors-7ba9f4d0c4cfcc3f | paper_workflow_reconstruction | sentence | paragraph | In classical hypothesis testing, one has access to i.i.d. data from some underlying distribution(s), and one wishes to test some property about them; this includes sequential tests dating back to Wald [18]. However, it is not often appreciated that it is possible to test hypotheses about a finite list of numbers that d... | [
{
"start": 0,
"end": 366,
"label": 0
},
{
"start": 366,
"end": 427,
"label": 1
},
{
"start": 427,
"end": 550,
"label": 0
}
] | 366 | 427 | Recalling the setup of Section 1.1, this is the nonstandard setting we find ourselves in. | In the setup of Section 1.1, we are in a nonstandard setting. | mixed | 1 | f4414fc243de0f9c21b383097a6cd6495653b868e90b94fcb91d4eb7752e6649 | NeurIPS | 2,020 | https://proceedings.neurips.cc/paper_files/paper/2020/file/e96c7de8f6390b1e6c71556e4e0a4959-Paper.pdf | 3a3c8cdb16186b6211faf5d287af949cfecab12ef3f57baaad9a634fb4fd5952 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Waudby-Smith, Ian",
"Ramdas, Aaditya"
] | Confidence sequences for sampling without replacement | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 7, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 7, "original_end": 8, "candidate_start": 1, "candidate_end": 2}, {"tag": "delete", "ori... |
icml-2017-d96a271d3a0abf3f/g01/sentence/context | icml-2017-d96a271d3a0abf3f | null | icml-2017-d96a271d3a0abf3f/g01 | authors-d0838d616b34d774 | paper_workflow_reconstruction | sentence | context | Line 7 of Algorithm 1 is known as the Schur update. Given a sparse input, this is the only step of the LU factorization that causes fill. As the algorithm progresses, fill will compound and may become dense, but the LU factorization, and truncated LU in particular, generally preserves some, if not most, of the sparsity... | [
{
"start": 0,
"end": 502,
"label": 0
},
{
"start": 502,
"end": 659,
"label": 1
},
{
"start": 659,
"end": 1672,
"label": 0
}
] | 502 | 659 | The column pivoting entails selecting the next b columns so that the in-place LU step is performed on a non-singular matrix (provided the remaining entries are not all zero). | When the remaining entries are not all zero, column pivoting chooses the next b columns so that the in-place LU step is carried out on a non-singular matrix. | mixed | 1 | beea1cf0c973a5c3ecabf0df21cb0e4121071aee3fea8c8c68fd58dbbdd96ae9 | ICML | 2,017 | https://proceedings.mlr.press/v70/anderson17a/anderson17a.pdf | 0879afc8a8e83176616b8e6643568fce7c8d1f10cf6b6d04d8b0a7b29df0a569 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"David Anderson",
"Ming Gu"
] | An Efficient, Sparsity-Preserving, Online Algorithm for Low-Rank Approximation | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 1, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 1, "original_end": 3, "candidate_start": 1, "candidate_end": 3}, {"tag": "insert", "ori... |
icml-2017-d96a271d3a0abf3f/g01/sentence/paragraph | icml-2017-d96a271d3a0abf3f | null | icml-2017-d96a271d3a0abf3f/g01 | authors-d0838d616b34d774 | paper_workflow_reconstruction | sentence | paragraph | There are many variations of the LU factorization. In Algorithm 2 the Crout version of LU is presented in block form. When the remaining entries are not all zero, column pivoting chooses the next b columns so that the in-place LU step is carried out on a non-singular matrix. Note that the matrix multiplication steps ar... | [
{
"start": 0,
"end": 118,
"label": 0
},
{
"start": 118,
"end": 275,
"label": 1
},
{
"start": 275,
"end": 400,
"label": 0
}
] | 118 | 275 | The column pivoting entails selecting the next b columns so that the in-place LU step is performed on a non-singular matrix (provided the remaining entries are not all zero). | When the remaining entries are not all zero, column pivoting chooses the next b columns so that the in-place LU step is carried out on a non-singular matrix. | mixed | 1 | bd597ac76152dc482e4b7a703cedfaa353d1c2e1002f3df9569ad5790ed55802 | ICML | 2,017 | https://proceedings.mlr.press/v70/anderson17a/anderson17a.pdf | 0879afc8a8e83176616b8e6643568fce7c8d1f10cf6b6d04d8b0a7b29df0a569 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"David Anderson",
"Ming Gu"
] | An Efficient, Sparsity-Preserving, Online Algorithm for Low-Rank Approximation | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 1, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 1, "original_end": 3, "candidate_start": 1, "candidate_end": 3}, {"tag": "insert", "ori... |
icml-2015-89b077cec0b2a1b9/g01/sentence/context | icml-2015-89b077cec0b2a1b9 | null | icml-2015-89b077cec0b2a1b9/g01 | authors-5f62b6ad7fea8bd5 | paper_workflow_reconstruction | sentence | context | To show the efficiency of our proposed LEML, we finally compare the training and testing time of the related SPD- based manifold/metric learning methods on YTC using an Intel(R) Core(TM) i7-37700M (3.40GHz) PC. The average training time and testing time for each method is tabulated in Table 2. Compared with ITML-LEM, o... | [
{
"start": 0,
"end": 1035,
"label": 0
},
{
"start": 1035,
"end": 1139,
"label": 1
},
{
"start": 1139,
"end": 1874,
"label": 0
}
] | 1,035 | 1,139 | The extensive experiments have shown that our proposed method is competitive to the state-of-the-art approaches. | Extensive experiments have shown our proposed method to be competitive with state-of-the-art approaches. | mixed | 1 | ae09660eab6911e1e9aa5e21d58c5c95b50acccff39006312c92757058069e12 | ICML | 2,015 | https://proceedings.mlr.press/v37/huanga15.pdf | 7643435ffb37f721be08fd7407ae2ffd74a1b33885f47c9090389ecf21470872 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Zhiwu Huang",
"Ruiping Wang",
"Shiguang Shan",
"Xianqiu Li",
"Xilin Chen"
] | Log-Euclidean Metric Learning on Symmetric Positive Definite Manifold with Application to Image Set Classification | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 5, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 5, "original_end": 37, "candidate_start": 1, "candidate_end": 33}, {"tag": "delete", "o... |
icml-2015-89b077cec0b2a1b9/g01/sentence/paragraph | icml-2015-89b077cec0b2a1b9 | null | icml-2015-89b077cec0b2a1b9/g01 | authors-5f62b6ad7fea8bd5 | paper_workflow_reconstruction | sentence | paragraph | In this paper we have proposed a novel metric learning framework on the SPD manifold for image set classification. Different from most existing methods, our approach seeks to learn the tangent map directly transforming the SPD matrix logarithms (rather than a vector) from the tangent space of the original SPD manifold ... | [
{
"start": 0,
"end": 344,
"label": 0
},
{
"start": 344,
"end": 448,
"label": 1
},
{
"start": 448,
"end": 633,
"label": 0
}
] | 344 | 448 | The extensive experiments have shown that our proposed method is competitive to the state-of-the-art approaches. | Extensive experiments have shown our proposed method to be competitive with state-of-the-art approaches. | mixed | 1 | fdeff22502dac9f50d79b20047981a7443ad9d443837573f9c644c6def4d9d67 | ICML | 2,015 | https://proceedings.mlr.press/v37/huanga15.pdf | 7643435ffb37f721be08fd7407ae2ffd74a1b33885f47c9090389ecf21470872 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Zhiwu Huang",
"Ruiping Wang",
"Shiguang Shan",
"Xianqiu Li",
"Xilin Chen"
] | Log-Euclidean Metric Learning on Symmetric Positive Definite Manifold with Application to Image Set Classification | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 5, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 5, "original_end": 37, "candidate_start": 1, "candidate_end": 33}, {"tag": "delete", "o... |
icml-2021-53267f81a5133d0e/g01/sentence/context | icml-2021-53267f81a5133d0e | null | icml-2021-53267f81a5133d0e/g01 | authors-950d07cdd00cbe99 | paper_workflow_reconstruction | sentence | context | The wrapped TFD distribution is identical to eq. (21) with ∆t pq as the time difference variable ∆t in the F 2 and F 3 components (eqs. (17) and (18) respectively).
In this section we evaluate the quality of pseudo-Riemannian embeddings via a series of graph reconstruction and link prediction experiments. Using small ... | [
{
"start": 0,
"end": 470,
"label": 0
},
{
"start": 470,
"end": 621,
"label": 1
},
{
"start": 621,
"end": 1009,
"label": 0
}
] | 470 | 621 | We then run the model on a series of benchmarks and an ablation study to characterize the utility of these embeddings in downstream applications. | We then run the model on a series of benchmarks and conduct an ablation study to characterize the utility of the embeddings in downstream applications. | mixed | 1 | 5b1b0ea464989c308e10c00ecde8e32537a20b98811ab6065168fb0ef04efda8 | ICML | 2,021 | https://proceedings.mlr.press/v139/sim21a/sim21a.pdf | 80c7b31a137ab6f44cafbafc999018d2da6175b74633f2134282f65aeb0e5099 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Aaron Sim",
"Maciej L Wiatrak",
"Angus Brayne",
"Paidi Creed",
"Saee Paliwal"
] | Directed Graph Embeddings in Pseudo-Riemannian Manifolds | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 51, "candidate_start": 0, "candidate_end": 51}, {"tag": "insert", "original_start": 51, "original_end": 51, "candidate_start": 51, "candidate_end": 59}, {"tag": "equal", ... |
icml-2021-53267f81a5133d0e/g01/sentence/paragraph | icml-2021-53267f81a5133d0e | null | icml-2021-53267f81a5133d0e/g01 | authors-950d07cdd00cbe99 | paper_workflow_reconstruction | sentence | paragraph | In this section we evaluate the quality of pseudo-Riemannian embeddings via a series of graph reconstruction and link prediction experiments. Using small synthetic graphs, we begin with a demonstration of the model’s ability to encode the particular set of graph-specific features outlined in Section 3. We then run the ... | [
{
"start": 0,
"end": 304,
"label": 0
},
{
"start": 304,
"end": 455,
"label": 1
},
{
"start": 455,
"end": 560,
"label": 0
}
] | 304 | 455 | We then run the model on a series of benchmarks and an ablation study to characterize the utility of these embeddings in downstream applications. | We then run the model on a series of benchmarks and conduct an ablation study to characterize the utility of the embeddings in downstream applications. | mixed | 1 | 41504a7488223c1ba4c9022cd9bbab5ab06b060ab81a9e1f17cbdac881518dac | ICML | 2,021 | https://proceedings.mlr.press/v139/sim21a/sim21a.pdf | 80c7b31a137ab6f44cafbafc999018d2da6175b74633f2134282f65aeb0e5099 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Aaron Sim",
"Maciej L Wiatrak",
"Angus Brayne",
"Paidi Creed",
"Saee Paliwal"
] | Directed Graph Embeddings in Pseudo-Riemannian Manifolds | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 51, "candidate_start": 0, "candidate_end": 51}, {"tag": "insert", "original_start": 51, "original_end": 51, "candidate_start": 51, "candidate_end": 59}, {"tag": "equal", ... |
9f44e956e3a2b7b5598c625fcc802c36/g01/sentence/context | 9f44e956e3a2b7b5598c625fcc802c36 | null | 9f44e956e3a2b7b5598c625fcc802c36/g01 | authors-e10e5ed36b9a89f3 | paper_workflow_reconstruction | sentence | context | We consider the problem of active sequential hypothesis testing where a Bayesian decision maker must infer the true hypothesis from a set of hypotheses. The decision maker may choose for a set of actions, where the outcome of an action is corrupted by independent noise. In this paper we consider a special case where th... | [
{
"start": 0,
"end": 1211,
"label": 0
},
{
"start": 1211,
"end": 1318,
"label": 1
},
{
"start": 1318,
"end": 2453,
"label": 0
}
] | 1,211 | 1,318 | In prior work, the probability distribution functions on the outcomes are assumed to be known. | In prior work, the probability distribution functions associated with the outcomes are assumed to be known. | mixed | 1 | 54cfcac881f6daf089a35726bd5cc5fead3a356d1845cdb5adc385e6384c0bee | NeurIPS | 2,017 | https://proceedings.neurips.cc/paper_files/paper/2017/file/9f44e956e3a2b7b5598c625fcc802c36-Paper.pdf | f6ed2693dae9f8c0cecbd54731f9b4f4b68c2777bd4fbf70957cfdd9b6ee7796 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Cecchi, Fabio",
"Hegde, Nidhi"
] | Adaptive Active Hypothesis Testing under Limited Information | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 54, "candidate_start": 0, "candidate_end": 54}, {"tag": "insert", "original_start": 54, "original_end": 54, "candidate_start": 54, "candidate_end": 57}, {"tag": "equal", ... |
9f44e956e3a2b7b5598c625fcc802c36/g01/sentence/paragraph | 9f44e956e3a2b7b5598c625fcc802c36 | null | 9f44e956e3a2b7b5598c625fcc802c36/g01 | authors-e10e5ed36b9a89f3 | paper_workflow_reconstruction | sentence | paragraph | We consider the problem of active sequential hypothesis testing with incomplete information. The original problem, first studied by Chernoff [1], is one where a Bayesian decision maker must infer the correct hypothesis from a set of J hypotheses. At each step the decision maker may choose from W actions where the outco... | [
{
"start": 0,
"end": 418,
"label": 0
},
{
"start": 418,
"end": 525,
"label": 1
},
{
"start": 525,
"end": 707,
"label": 0
}
] | 418 | 525 | In prior work, the probability distribution functions on the outcomes are assumed to be known. | In prior work, the probability distribution functions associated with the outcomes are assumed to be known. | mixed | 1 | c01a396996ecfff72cbb9621ae4719db13847e26a615c0e6e132c2b3079fddf5 | NeurIPS | 2,017 | https://proceedings.neurips.cc/paper_files/paper/2017/file/9f44e956e3a2b7b5598c625fcc802c36-Paper.pdf | f6ed2693dae9f8c0cecbd54731f9b4f4b68c2777bd4fbf70957cfdd9b6ee7796 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Cecchi, Fabio",
"Hegde, Nidhi"
] | Adaptive Active Hypothesis Testing under Limited Information | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 54, "candidate_start": 0, "candidate_end": 54}, {"tag": "insert", "original_start": 54, "original_end": 54, "candidate_start": 54, "candidate_end": 57}, {"tag": "equal", ... |
icml-2016-3f46a6be26b96414/g01/sentence/context | icml-2016-3f46a6be26b96414 | null | icml-2016-3f46a6be26b96414/g01 | authors-8af4fd9ce740f2ed | paper_workflow_reconstruction | sentence | context | The above proposition shows that one step of Newton method for analytic center requires 1 dense matrix multi-plication and solving 1 linear system; and for volumetric center, it requires 5 dense matrix multiplications, 1 matrix inversion and solving 1 linear system if implemented correctly. Although the analytic center... | [
{
"start": 0,
"end": 1202,
"label": 0
},
{
"start": 1202,
"end": 1454,
"label": 1
},
{
"start": 1454,
"end": 2235,
"label": 0
}
] | 1,202 | 1,454 | To compensate on the unfairness to TFOCS, we note that the algorithm TFOCS uses is accelerated gradient descent and hence we implement the Gonzaga-Karas’s accelerated gradient descent (Gonzaga and Karas, 2013), which is specifically designed to be used with exact line search. | To compensate for the unfairness toward TFOCS, we note that it uses accelerated gradient descent; hence, we implement Gonzaga-Karas’s accelerated gradient descent (Gonzaga and Karas, 2013), which is specifically designed for use with exact line search. | mixed | 1 | adfdd8383f9666cd342b9cf36103dcf058a411b8478c6bc6cfc21df6930ca4e2 | ICML | 2,016 | https://proceedings.mlr.press/v48/bubeck16.pdf | 432947250ff82b007ffe56e611c523a52843d5f9a526e8f5500be0e98ec4ae3a | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Sebastien Bubeck",
"Yin Tat Lee"
] | Black-box Optimization with a Politician | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 14, "candidate_start": 0, "candidate_end": 14}, {"tag": "insert", "original_start": 14, "original_end": 14, "candidate_start": 14, "candidate_end": 15}, {"tag": "equal", ... |
icml-2016-3f46a6be26b96414/g01/sentence/paragraph | icml-2016-3f46a6be26b96414 | null | icml-2016-3f46a6be26b96414/g01 | authors-8af4fd9ce740f2ed | paper_workflow_reconstruction | sentence | paragraph | In this section, we compare the geometric politician against two libraries for first order methods, minFunc (Schmidt, 2012) and TFOCS (Becker et al., 2011). Both are popular MATLAB libraries for minimizing general smooth convex functions. Since the focus of this paper is all about how to find a good step direction usin... | [
{
"start": 0,
"end": 624,
"label": 0
},
{
"start": 624,
"end": 876,
"label": 1
},
{
"start": 876,
"end": 1165,
"label": 0
}
] | 624 | 876 | To compensate on the unfairness to TFOCS, we note that the algorithm TFOCS uses is accelerated gradient descent and hence we implement the Gonzaga-Karas’s accelerated gradient descent (Gonzaga and Karas, 2013), which is specifically designed to be used with exact line search. | To compensate for the unfairness toward TFOCS, we note that it uses accelerated gradient descent; hence, we implement Gonzaga-Karas’s accelerated gradient descent (Gonzaga and Karas, 2013), which is specifically designed for use with exact line search. | mixed | 1 | 992816136aa6764ebe996c28425e8053d3d4a85578d3306dbbfb953104b2e25c | ICML | 2,016 | https://proceedings.mlr.press/v48/bubeck16.pdf | 432947250ff82b007ffe56e611c523a52843d5f9a526e8f5500be0e98ec4ae3a | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Sebastien Bubeck",
"Yin Tat Lee"
] | Black-box Optimization with a Politician | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 14, "candidate_start": 0, "candidate_end": 14}, {"tag": "insert", "original_start": 14, "original_end": 14, "candidate_start": 14, "candidate_end": 15}, {"tag": "equal", ... |
icml-2015-cc134b564b2accc6/g01/sentence/context | icml-2015-cc134b564b2accc6 | null | icml-2015-cc134b564b2accc6/g01 | authors-05dcfb48249aa175 | paper_workflow_reconstruction | sentence | context | Solution path algorithms have been proposed for several important learning algorithms. For example, Hastie et al. (2004) proposed a solution path algorithm for C-support vector classification (C-SVC). Bach et al. (2006) proposed solution path algorithm for 2C-support vector classification (2C-SVC). Gunter & Zhu (2007),... | [
{
"start": 0,
"end": 1307,
"label": 0
},
{
"start": 1307,
"end": 1398,
"label": 1
},
{
"start": 1398,
"end": 3430,
"label": 0
}
] | 1,307 | 1,398 | That is, previous solution path algorithms do not directly lead to global minimal CV error. | That is, previous solution path algorithms do not directly lead to global minimal CV error. | mixed | 1 | e0432a8a78c3c173831adb8a7de58cce0338b58f0f47d7a76195d358c2f5feb1 | ICML | 2,015 | https://proceedings.mlr.press/v37/gu15.pdf | e9d6170bc9ada17dca49a9ca545e4b7b4869b349709d9262f712f13f82b4d34a | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Bin Gu",
"Charles Ling"
] | A New Generalized Error Path Algorithm for Model Selection | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 91, "candidate_start": 0, "candidate_end": 91}], "sentence_boundary_method": "quote/parenthesis/citation-aware heuristic v2, not gold syntax", "similarity": {"word_lcs_f1... |
icml-2015-cc134b564b2accc6/g01/sentence/paragraph | icml-2015-cc134b564b2accc6 | null | icml-2015-cc134b564b2accc6/g01 | authors-05dcfb48249aa175 | paper_workflow_reconstruction | sentence | paragraph | Solution path algorithms can fit the entire solutions with respect to one parameter. However, they still do not guarantee to find the model with minimum CV error, as they do not correspond to “error path” in CV. That is, previous solution path algorithms do not directly lead to global minimal CV error. To the best of o... | [
{
"start": 0,
"end": 212,
"label": 0
},
{
"start": 212,
"end": 303,
"label": 1
},
{
"start": 303,
"end": 977,
"label": 0
}
] | 212 | 303 | That is, previous solution path algorithms do not directly lead to global minimal CV error. | That is, previous solution path algorithms do not directly lead to global minimal CV error. | mixed | 1 | 2ad15cf48edb72d54f3cc997e770a3868cc6a433f8ef838661b077e249e94509 | ICML | 2,015 | https://proceedings.mlr.press/v37/gu15.pdf | e9d6170bc9ada17dca49a9ca545e4b7b4869b349709d9262f712f13f82b4d34a | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Bin Gu",
"Charles Ling"
] | A New Generalized Error Path Algorithm for Model Selection | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 91, "candidate_start": 0, "candidate_end": 91}], "sentence_boundary_method": "quote/parenthesis/citation-aware heuristic v2, not gold syntax", "similarity": {"word_lcs_f1... |
30ef30b64204a3088a26bc2e6ecf7602/g01/sentence/context | 30ef30b64204a3088a26bc2e6ecf7602 | null | 30ef30b64204a3088a26bc2e6ecf7602/g01 | authors-1e6cec6326ae3cb7 | paper_workflow_reconstruction | sentence | context | In this paper, our focus is on semi-supervised deep learning. There has always been interest in exploiting unlabeled data to improve the performance of ConvNets. One approach is to use unlabeled data to pre-train the filters of ConvNet [18, 19]. The goal is to reduce the number of training epochs required to converge a... | [
{
"start": 0,
"end": 1639,
"label": 0
},
{
"start": 1639,
"end": 1851,
"label": 1
},
{
"start": 1851,
"end": 3139,
"label": 0
}
] | 1,639 | 1,851 | Another notable example is semi-supervised learning with ladder networks [28] in which the sums of supervised and unsupervised loss functions are simultaneously minimized by backpropagation. | Another notable example of semi-supervised learning is ladder networks [28], which combine a supervised loss function and an unsupervised loss function whose sum is simultaneously minimized using backpropagation. | mixed | 1 | 5faebf662f3337a7e99f30dbb9913c3e862c8373b23d01d29c525630caed5a31 | NeurIPS | 2,016 | https://proceedings.neurips.cc/paper_files/paper/2016/file/30ef30b64204a3088a26bc2e6ecf7602-Paper.pdf | 5ac4f1ce2fa249be5b8d8f5cf6dfdb90e4973df19e30abbd3a401efaa3b8961b | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Sajjadi, Mehdi",
"Javanmardi, Mehran",
"Tasdizen, Tolga"
] | Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 24, "candidate_start": 0, "candidate_end": 24}, {"tag": "replace", "original_start": 24, "original_end": 26, "candidate_start": 24, "candidate_end": 26}, {"tag": "equal",... |
30ef30b64204a3088a26bc2e6ecf7602/g01/sentence/paragraph | 30ef30b64204a3088a26bc2e6ecf7602 | null | 30ef30b64204a3088a26bc2e6ecf7602/g01 | authors-1e6cec6326ae3cb7 | paper_workflow_reconstruction | sentence | paragraph | Another example of semi-supervised learning with ConvNets is region embedding [24], which is used for text categorization. The work in [25] is also a deep semi-supervised learning method based on embedding techniques. Unlabeled video frames are also being used to train ConvNets [26, 27]. The target of the ConvNet is ca... | [
{
"start": 0,
"end": 377,
"label": 0
},
{
"start": 377,
"end": 589,
"label": 1
},
{
"start": 589,
"end": 1006,
"label": 0
}
] | 377 | 589 | Another notable example is semi-supervised learning with ladder networks [28] in which the sums of supervised and unsupervised loss functions are simultaneously minimized by backpropagation. | Another notable example of semi-supervised learning is ladder networks [28], which combine a supervised loss function and an unsupervised loss function whose sum is simultaneously minimized using backpropagation. | mixed | 1 | 917f59486e2e3d359a969034301b3240e98fdbb5e7d862a2ba34e08c97678e70 | NeurIPS | 2,016 | https://proceedings.neurips.cc/paper_files/paper/2016/file/30ef30b64204a3088a26bc2e6ecf7602-Paper.pdf | 5ac4f1ce2fa249be5b8d8f5cf6dfdb90e4973df19e30abbd3a401efaa3b8961b | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Sajjadi, Mehdi",
"Javanmardi, Mehran",
"Tasdizen, Tolga"
] | Regularization With Stochastic Transformations and Perturbations for Deep Semi-Supervised Learning | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 24, "candidate_start": 0, "candidate_end": 24}, {"tag": "replace", "original_start": 24, "original_end": 26, "candidate_start": 24, "candidate_end": 26}, {"tag": "equal",... |
9af08cda54faea9adf40a201794183cf/g01/sentence/context | 9af08cda54faea9adf40a201794183cf | null | 9af08cda54faea9adf40a201794183cf/g01 | authors-b93b92d2f8b7bda4 | paper_workflow_reconstruction | sentence | context | Last Layer Gradients: Deep models have numerous parameters leading to very high dimensional gradients. Since our kernel matrix is computed using the cosine similarity of gradients, this becomes intractable for most models. To solve this problem, we use last-layer gradient approximation by representing data points using... | [
{
"start": 0,
"end": 908,
"label": 0
},
{
"start": 908,
"end": 1134,
"label": 1
},
{
"start": 1134,
"end": 2280,
"label": 0
}
] | 908 | 1,134 | Using these experiments, we cover the issues with the current AL methods and show that these issues can be mitigated by using a unified implementation using SCMI with appropriate choices of query and/or conditioning sets. | Based on these experiments, we address issues associated with current AL methods and show that they can be mitigated through a unified implementation using SCMI, with appropriate choices of query sets and/or conditioning sets. | mixed | 1 | 23f4ea80dbc512593ea779c71a7bcf86cd773b28e83a8e15d76c6d920709ec00 | NeurIPS | 2,021 | https://proceedings.neurips.cc/paper_files/paper/2021/file/9af08cda54faea9adf40a201794183cf-Paper.pdf | 1bf9691093f1c71ecdcf323d9f5b0c5df5ac7704fbb3d33c7425f9d739647fdc | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Kothawade, Suraj",
"Beck, Nathan",
"Killamsetty, Krishnateja",
"Iyer, Rishabh"
] | SIMILAR: Submodular Information Measures Based Active Learning In Realistic Scenarios | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 1, "candidate_start": 0, "candidate_end": 2}, {"tag": "equal", "original_start": 1, "original_end": 2, "candidate_start": 2, "candidate_end": 3}, {"tag": "replace", "or... |
9af08cda54faea9adf40a201794183cf/g01/sentence/paragraph | 9af08cda54faea9adf40a201794183cf | null | 9af08cda54faea9adf40a201794183cf/g01 | authors-b93b92d2f8b7bda4 | paper_workflow_reconstruction | sentence | paragraph | In this section, we empirically evaluate the effectiveness of S IMILAR on a wide range of scenarios like rare classes (Sec. 4.1), redundancy (Sec. 4.2) and out-of-distribution (Sec. 4.3). We do so by comparing the accuracy and selections of various SCMI based acquisition functions with existing AL approaches. Based on ... | [
{
"start": 0,
"end": 311,
"label": 0
},
{
"start": 311,
"end": 537,
"label": 1
},
{
"start": 537,
"end": 909,
"label": 0
}
] | 311 | 537 | Using these experiments, we cover the issues with the current AL methods and show that these issues can be mitigated by using a unified implementation using SCMI with appropriate choices of query and/or conditioning sets. | Based on these experiments, we address issues associated with current AL methods and show that they can be mitigated through a unified implementation using SCMI, with appropriate choices of query sets and/or conditioning sets. | mixed | 1 | 8a8613c5cc89636d1834391825cfc7a0ed54df57381dd6a1731e5043906570f3 | NeurIPS | 2,021 | https://proceedings.neurips.cc/paper_files/paper/2021/file/9af08cda54faea9adf40a201794183cf-Paper.pdf | 1bf9691093f1c71ecdcf323d9f5b0c5df5ac7704fbb3d33c7425f9d739647fdc | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Kothawade, Suraj",
"Beck, Nathan",
"Killamsetty, Krishnateja",
"Iyer, Rishabh"
] | SIMILAR: Submodular Information Measures Based Active Learning In Realistic Scenarios | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 1, "candidate_start": 0, "candidate_end": 2}, {"tag": "equal", "original_start": 1, "original_end": 2, "candidate_start": 2, "candidate_end": 3}, {"tag": "replace", "or... |
icml-2019-6f44da33f738a5d0/g01/sentence/context | icml-2019-6f44da33f738a5d0 | null | icml-2019-6f44da33f738a5d0/g01 | authors-406a81ff58c8d16b | paper_workflow_reconstruction | sentence | context | In this manuscript, we show that K 33 -free zero-field Ising models are tractable in terms of inference and sampling 3 and give a tight asymptotic bound, O(N 2 ), for both operations. For that purpose, we use graph decomposition into triconnected components—the result of recursive splitting by pairs of vertices, discon... | [
{
"start": 0,
"end": 1465,
"label": 0
},
{
"start": 1465,
"end": 1637,
"label": 1
},
{
"start": 1637,
"end": 2301,
"label": 0
}
] | 1,465 | 1,637 | Section 5 discusses a scheme for polynomial inference and sampling in zero-field models over graphs with triconnected components that are either planar or of O(1) size. | Section 5 discusses a scheme for polynomial-time inference and sampling in zero-field models on graphs whose triconnected components are each either planar or of O(1) size. | mixed | 1 | 2b120c3e3f77a61e3bc552f2c04f93f31fb3a93e4f6aa7cf0538bbba665e4d1c | ICML | 2,019 | https://proceedings.mlr.press/v97/likhosherstov19a/likhosherstov19a.pdf | 8f847f1c508c18fc0657c19ee43d65f02cbaf83e21227f040cf2bc8d31382363 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Valerii Likhosherstov",
"Yury Maximov",
"Misha Chertkov"
] | Inference and Sampling of $K_33$-free Ising Models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 43, "candidate_start": 0, "candidate_end": 43}, {"tag": "insert", "original_start": 43, "original_end": 43, "candidate_start": 43, "candidate_end": 48}, {"tag": "equal", ... |
icml-2019-6f44da33f738a5d0/g01/sentence/paragraph | icml-2019-6f44da33f738a5d0 | null | icml-2019-6f44da33f738a5d0/g01 | authors-406a81ff58c8d16b | paper_workflow_reconstruction | sentence | paragraph | Structure. The manuscript is organized as follows. Sections 2 and 3, respectively, establish notations and pose problems of inference and sampling. Section 4 presents transition from the zero-field Ising model to an equivalent tractable perfect matching (PM) model. This provides a 3 description of a O(N 2 ) inference a... | [
{
"start": 0,
"end": 442,
"label": 0
},
{
"start": 442,
"end": 614,
"label": 1
},
{
"start": 614,
"end": 974,
"label": 0
}
] | 442 | 614 | Section 5 discusses a scheme for polynomial inference and sampling in zero-field models over graphs with triconnected components that are either planar or of O(1) size. | Section 5 discusses a scheme for polynomial-time inference and sampling in zero-field models on graphs whose triconnected components are each either planar or of O(1) size. | mixed | 1 | 2646f94d4b6f394043aab208d58ed2d59a35a2e9ea5d01fa7363964e6c64ca2d | ICML | 2,019 | https://proceedings.mlr.press/v97/likhosherstov19a/likhosherstov19a.pdf | 8f847f1c508c18fc0657c19ee43d65f02cbaf83e21227f040cf2bc8d31382363 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Valerii Likhosherstov",
"Yury Maximov",
"Misha Chertkov"
] | Inference and Sampling of $K_33$-free Ising Models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 43, "candidate_start": 0, "candidate_end": 43}, {"tag": "insert", "original_start": 43, "original_end": 43, "candidate_start": 43, "candidate_end": 48}, {"tag": "equal", ... |
acl-2014-dda43cf2886bb0c6/g01/sentence/context | acl-2014-dda43cf2886bb0c6 | null | acl-2014-dda43cf2886bb0c6/g01 | authors-9a911ffa90d46935 | paper_workflow_reconstruction | sentence | context | For YA, we used the standard implementations for P@1 and mean reciprocal rank (MRR) (Manning et al., 2008). In the Bio corpus, because answer candidates are not guaranteed to match gold annotations exactly, these metrics do not immediately apply. We adapted them to this dataset by weighing each answer by its overlap wi... | [
{
"start": 0,
"end": 886,
"label": 0
},
{
"start": 886,
"end": 1031,
"label": 1
},
{
"start": 1031,
"end": 1348,
"label": 0
}
] | 886 | 1,031 | For example, SVMs with polynomial kernels of degree two showed approximately half a percent (absolute) performance gain over the linear kernel. | For example, SVMs using degree-two polynomial kernels showed an absolute performance gain of approximately half a percent over the linear kernel. | mixed | 1 | de0c2a3c5260774c8f2c4622a9326859ef7fe8295c9719547524a433c8087f01 | ACL | 2,014 | https://aclanthology.org/P14-1092.pdf | 171dbbc3afb9fe3a29f59900584020fb001aa965573bb9c89ad934bf2672fdc9 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Peter Jansen",
"Mihai Surdeanu",
"Peter Clark"
] | Discourse Complements Lexical Semantics for Non-factoid Answer Reranking | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 18, "candidate_start": 0, "candidate_end": 18}, {"tag": "insert", "original_start": 18, "original_end": 18, "candidate_start": 18, "candidate_end": 32}, {"tag": "equal", ... |
acl-2014-dda43cf2886bb0c6/g01/sentence/paragraph | acl-2014-dda43cf2886bb0c6 | null | acl-2014-dda43cf2886bb0c6/g01 | authors-9a911ffa90d46935 | paper_workflow_reconstruction | sentence | paragraph | The performance of all models can ultimately be increased by using more sophisticated learning frameworks, and considering more answer candidates in CR (for Bio). For example, SVMs using degree-two polynomial kernels showed an absolute performance gain of approximately half a percent over the linear kernel. However, th... | [
{
"start": 0,
"end": 163,
"label": 0
},
{
"start": 163,
"end": 308,
"label": 1
},
{
"start": 308,
"end": 497,
"label": 0
}
] | 163 | 308 | For example, SVMs with polynomial kernels of degree two showed approximately half a percent (absolute) performance gain over the linear kernel. | For example, SVMs using degree-two polynomial kernels showed an absolute performance gain of approximately half a percent over the linear kernel. | mixed | 1 | 9bd1c644fa1b009ad666cf6169db6b8f02210550216a9ae8b6a3543563e97420 | ACL | 2,014 | https://aclanthology.org/P14-1092.pdf | 171dbbc3afb9fe3a29f59900584020fb001aa965573bb9c89ad934bf2672fdc9 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Peter Jansen",
"Mihai Surdeanu",
"Peter Clark"
] | Discourse Complements Lexical Semantics for Non-factoid Answer Reranking | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 18, "candidate_start": 0, "candidate_end": 18}, {"tag": "insert", "original_start": 18, "original_end": 18, "candidate_start": 18, "candidate_end": 32}, {"tag": "equal", ... |
icml-2018-3d80bff1abcc4137/g01/sentence/context | icml-2018-3d80bff1abcc4137 | null | icml-2018-3d80bff1abcc4137/g01 | authors-82f94c00f71bfd76 | paper_workflow_reconstruction | sentence | context | Note that with this approach once the MST is constructed it can be used to compute k-SLC for any value of k. Furthermore, it induces a hierarchical clustering structure that is often desirable in practice. According to Manning, Raghavan and Schütze (Manning et al., 2008) the main impediment to this approach in practice... | [
{
"start": 0,
"end": 1236,
"label": 0
},
{
"start": 1236,
"end": 1469,
"label": 1
},
{
"start": 1469,
"end": 2340,
"label": 0
}
] | 1,236 | 1,469 | It has emerged through a sequence of papers (Feldman et al., 2008; Karloff et al., 2010; Goodrich et al., 2011; Beame et al., 2013; Andoni et al., 2014) and has been analyzed extensively (Fish et al., 2015; Roughgarden et al., 2016). | It has emerged through a sequence of papers (Feldman et al., 2008; Karloff et al., 2010; Goodrich et al., 2011; Beame et al., 2013; Andoni et al., 2014) and has been extensively analyzed (Fish et al., 2015; Roughgarden et al., 2016). | mixed | 1 | fab93a0a810a55927a78ac7eb3b78eb0a99fa1ab7c7ced3e749c2ebb0fb22a26 | ICML | 2,018 | https://proceedings.mlr.press/v80/yaroslavtsev18a/yaroslavtsev18a.pdf | b38e13426ce1c3f17ef9ec0fc81fcfae6418630fae20924aabd753a1a7083dad | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Grigory Yaroslavtsev",
"Adithya Vadapalli"
] | Massively Parallel Algorithms and Hardness for Single-Linkage Clustering under $\ell_p$ Distances | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 166, "candidate_start": 0, "candidate_end": 166}, {"tag": "delete", "original_start": 166, "original_end": 175, "candidate_start": 166, "candidate_end": 166}, {"tag": "eq... |
icml-2018-3d80bff1abcc4137/g01/sentence/paragraph | icml-2018-3d80bff1abcc4137 | null | icml-2018-3d80bff1abcc4137/g01 | authors-82f94c00f71bfd76 | paper_workflow_reconstruction | sentence | paragraph | We present analysis of performance of our algorithms in the Massively Parallel Computation model (MPC) which is the most commonly used theoretical model of computation on synchronous large-scale data processing platforms such as MapReduce and Spark. As we demonstrate through experiments in Spark this model accurately r... | [
{
"start": 0,
"end": 424,
"label": 0
},
{
"start": 424,
"end": 657,
"label": 1
},
{
"start": 657,
"end": 860,
"label": 0
}
] | 424 | 657 | It has emerged through a sequence of papers (Feldman et al., 2008; Karloff et al., 2010; Goodrich et al., 2011; Beame et al., 2013; Andoni et al., 2014) and has been analyzed extensively (Fish et al., 2015; Roughgarden et al., 2016). | It has emerged through a sequence of papers (Feldman et al., 2008; Karloff et al., 2010; Goodrich et al., 2011; Beame et al., 2013; Andoni et al., 2014) and has been extensively analyzed (Fish et al., 2015; Roughgarden et al., 2016). | mixed | 1 | aadb370d068b76906eaea681149d3b10d5ee48470e0881b26a9ce4cb14ff85c0 | ICML | 2,018 | https://proceedings.mlr.press/v80/yaroslavtsev18a/yaroslavtsev18a.pdf | b38e13426ce1c3f17ef9ec0fc81fcfae6418630fae20924aabd753a1a7083dad | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Grigory Yaroslavtsev",
"Adithya Vadapalli"
] | Massively Parallel Algorithms and Hardness for Single-Linkage Clustering under $\ell_p$ Distances | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 166, "candidate_start": 0, "candidate_end": 166}, {"tag": "delete", "original_start": 166, "original_end": 175, "candidate_start": 166, "candidate_end": 166}, {"tag": "eq... |
4d95d05a4fc4eadbc3b9dde67afdca39/g01/sentence/context | 4d95d05a4fc4eadbc3b9dde67afdca39 | null | 4d95d05a4fc4eadbc3b9dde67afdca39/g01 | authors-73143df6f7a1f7de | paper_workflow_reconstruction | sentence | context | We also obtain analogous worst and average-case l_1 bounds with scaling factors on the same order. Due to space constraints, please see Theorem 5 and 6 in the appendix for the results.
Lastly, we note that these bounds are general. In the appendix, we obtain a simple derivation of the CGA-specific bias, which can be p... | [
{
"start": 0,
"end": 362,
"label": 0
},
{
"start": 362,
"end": 522,
"label": 1
},
{
"start": 522,
"end": 1392,
"label": 0
}
] | 362 | 522 | Note that bias in the estimation of the CF only arises due to model misspecification, i.e if order k is used to model a game of order r for r k. | Note that bias in estimating the CF arises only from model misspecification, which is using order k to model a game of order r under the stated condition “r k.” | mixed | 1 | 106bdf7de44452faef3954e25b738b39d0f980308744b0d6c2616a28c9d32032 | NeurIPS | 2,020 | https://proceedings.neurips.cc/paper_files/paper/2020/file/4d95d05a4fc4eadbc3b9dde67afdca39-Paper.pdf | 49353a6c21107834672b7d4d546aa861a3f6da1abba0c3e0d2d8f55e40853b1c | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Yan, Tom",
"Kroer, Christian",
"Peysakhovich, Alexander"
] | Evaluating and Rewarding Teamwork Using Cooperative Game Abstractions | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 18, "candidate_start": 0, "candidate_end": 18}, {"tag": "delete", "original_start": 18, "original_end": 22, "candidate_start": 18, "candidate_end": 18}, {"tag": "equal", ... |
4d95d05a4fc4eadbc3b9dde67afdca39/g01/sentence/paragraph | 4d95d05a4fc4eadbc3b9dde67afdca39 | null | 4d95d05a4fc4eadbc3b9dde67afdca39/g01 | authors-73143df6f7a1f7de | paper_workflow_reconstruction | sentence | paragraph | Lastly, we note that these bounds are general. In the appendix, we obtain a simple derivation of the CGA-specific bias, which can be plugged into these bounds for the SV bias. Note that bias in estimating the CF arises only from model misspecification, which is using order k to model a game of order r under the stated ... | [
{
"start": 0,
"end": 176,
"label": 0
},
{
"start": 176,
"end": 336,
"label": 1
},
{
"start": 336,
"end": 619,
"label": 0
}
] | 176 | 336 | Note that bias in the estimation of the CF only arises due to model misspecification, i.e if order k is used to model a game of order r for r k. | Note that bias in estimating the CF arises only from model misspecification, which is using order k to model a game of order r under the stated condition “r k.” | mixed | 1 | b6086af2caca093ca300fb61cd13f645abbec2806ae432e18d9fc24a683c77f4 | NeurIPS | 2,020 | https://proceedings.neurips.cc/paper_files/paper/2020/file/4d95d05a4fc4eadbc3b9dde67afdca39-Paper.pdf | 49353a6c21107834672b7d4d546aa861a3f6da1abba0c3e0d2d8f55e40853b1c | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Yan, Tom",
"Kroer, Christian",
"Peysakhovich, Alexander"
] | Evaluating and Rewarding Teamwork Using Cooperative Game Abstractions | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 18, "candidate_start": 0, "candidate_end": 18}, {"tag": "delete", "original_start": 18, "original_end": 22, "candidate_start": 18, "candidate_end": 18}, {"tag": "equal", ... |
1cdf14d1e3699d61d237cf76ce1c2dca/g01/sentence/context | 1cdf14d1e3699d61d237cf76ce1c2dca | null | 1cdf14d1e3699d61d237cf76ce1c2dca/g01 | authors-b6e752e52241f818 | paper_workflow_reconstruction | sentence | context | Autoregressive Models AR models represent a distribution as a product of conditional, learnable factors via the chain rule of probability densities. While this makes them powerful models for density estimation [70, 24], their samples often lack global consistency. Especially on image data modeled with convolutional arc... | [
{
"start": 0,
"end": 1654,
"label": 0
},
{
"start": 1654,
"end": 1764,
"label": 1
},
{
"start": 1764,
"end": 2962,
"label": 0
}
] | 1,654 | 1,764 | However, AR modeling of a learned representation still limits applications compared to latent variable models. | However, AR modeling of a learned representation still limits applications relative to latent variable models. | mixed | 1 | febf9228c78a747e71a0742a55896cf81b1a3a4386bf9c48fc573e8c0ce0109a | NeurIPS | 2,021 | https://proceedings.neurips.cc/paper_files/paper/2021/file/1cdf14d1e3699d61d237cf76ce1c2dca-Paper.pdf | d073752bc021bae682f2bd40ffecff1b9964199f2d1e6ea74998a09ab68fdbc7 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Esser, Patrick",
"Rombach, Robin",
"Blattmann, Andreas",
"Ommer, Bjorn"
] | ImageBART: Bidirectional Context with Multinomial Diffusion for Autoregressive Image Synthesis | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 75, "candidate_start": 0, "candidate_end": 75}, {"tag": "delete", "original_start": 75, "original_end": 80, "candidate_start": 75, "candidate_end": 75}, {"tag": "equal", ... |
1cdf14d1e3699d61d237cf76ce1c2dca/g01/sentence/paragraph | 1cdf14d1e3699d61d237cf76ce1c2dca | null | 1cdf14d1e3699d61d237cf76ce1c2dca/g01 | authors-b6e752e52241f818 | paper_workflow_reconstruction | sentence | paragraph | Generative Models on Improved Representations Another successful line of work first learn an improved image representation and subsequently learn a generative model for this representation [74, 12]. Most works [58, 21, 56] learn a discrete representation which is subsequently modeled autoregressively but approaches usi... | [
{
"start": 0,
"end": 768,
"label": 0
},
{
"start": 768,
"end": 878,
"label": 1
},
{
"start": 878,
"end": 1097,
"label": 0
}
] | 768 | 878 | However, AR modeling of a learned representation still limits applications compared to latent variable models. | However, AR modeling of a learned representation still limits applications relative to latent variable models. | mixed | 1 | 366f8e620434754d06952949ea7af5f5a7d6d0a05ee461b6e72dc6ba53450994 | NeurIPS | 2,021 | https://proceedings.neurips.cc/paper_files/paper/2021/file/1cdf14d1e3699d61d237cf76ce1c2dca-Paper.pdf | d073752bc021bae682f2bd40ffecff1b9964199f2d1e6ea74998a09ab68fdbc7 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Esser, Patrick",
"Rombach, Robin",
"Blattmann, Andreas",
"Ommer, Bjorn"
] | ImageBART: Bidirectional Context with Multinomial Diffusion for Autoregressive Image Synthesis | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 75, "candidate_start": 0, "candidate_end": 75}, {"tag": "delete", "original_start": 75, "original_end": 80, "candidate_start": 75, "candidate_end": 75}, {"tag": "equal", ... |
acl-2020-025e49fc744bbe7c/g01/sentence/context | acl-2020-025e49fc744bbe7c | null | acl-2020-025e49fc744bbe7c/g01 | authors-13e8bfe132a93ebe | paper_workflow_reconstruction | sentence | context | For reasons of space, we display only subject task results, in Figure 4. All results may be found in our GitHub repository linked in Footnote 1.
Interim summary GPT shows uniform strong encoding of subject information and solid encoding of verb information on the target and subsequent words—but weak encoding of object... | [
{
"start": 0,
"end": 353,
"label": 0
},
{
"start": 353,
"end": 582,
"label": 1
},
{
"start": 582,
"end": 2408,
"label": 0
}
] | 353 | 582 | BERT and ELMo show more nuance in their distribution of the information types, with BERT heavily deprioritizing gender information, but strongly encoding animacy and maintaining rich number information for both nouns across all words. | BERT and ELMo exhibit greater nuance in how information types are distributed; BERT heavily deprioritizes gender information while strongly encoding animacy and maintaining rich number information for both nouns across all words. | mixed | 1 | 343b78697a93e7e2ea6d3de459aeae023f5cddb653808c81cfed3f2685188cc9 | ACL | 2,020 | https://aclanthology.org/2020.acl-main.434.pdf | c873bf0f043f0c69b6407e3489420e594b2d899ab82d566e4a2065e86a651312 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Josef Klafka",
"Allyson Ettinger"
] | Spying on Your Neighbors: Fine-grained Probing of Contextual Embeddings for Information about Surrounding Words | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 14, "candidate_start": 0, "candidate_end": 14}, {"tag": "replace", "original_start": 14, "original_end": 15, "candidate_start": 14, "candidate_end": 16}, {"tag": "equal",... |
acl-2020-025e49fc744bbe7c/g01/sentence/paragraph | acl-2020-025e49fc744bbe7c | null | acl-2020-025e49fc744bbe7c/g01 | authors-13e8bfe132a93ebe | paper_workflow_reconstruction | sentence | paragraph | Interim summary GPT shows uniform strong encoding of subject information and solid encoding of verb information on the target and subsequent words—but weak encoding of object information on the object noun. BERT and ELMo exhibit greater nuance in how information types are distributed; BERT heavily deprioritizes gender ... | [
{
"start": 0,
"end": 207,
"label": 0
},
{
"start": 207,
"end": 436,
"label": 1
},
{
"start": 436,
"end": 900,
"label": 0
}
] | 207 | 436 | BERT and ELMo show more nuance in their distribution of the information types, with BERT heavily deprioritizing gender information, but strongly encoding animacy and maintaining rich number information for both nouns across all words. | BERT and ELMo exhibit greater nuance in how information types are distributed; BERT heavily deprioritizes gender information while strongly encoding animacy and maintaining rich number information for both nouns across all words. | mixed | 1 | 430f35ec3e824a1799dfeb7eaa081c9f6dd05d0845b33df20c87305f178f11fc | ACL | 2,020 | https://aclanthology.org/2020.acl-main.434.pdf | c873bf0f043f0c69b6407e3489420e594b2d899ab82d566e4a2065e86a651312 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Josef Klafka",
"Allyson Ettinger"
] | Spying on Your Neighbors: Fine-grained Probing of Contextual Embeddings for Information about Surrounding Words | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 14, "candidate_start": 0, "candidate_end": 14}, {"tag": "replace", "original_start": 14, "original_end": 15, "candidate_start": 14, "candidate_end": 16}, {"tag": "equal",... |
b056eb1587586b71e2da9acfe4fbd19e/g01/sentence/context | b056eb1587586b71e2da9acfe4fbd19e | null | b056eb1587586b71e2da9acfe4fbd19e/g01 | authors-3952694096d86f7a | paper_workflow_reconstruction | sentence | context | EDML is a recently proposed algorithm for learning MAP parameters of a Bayesian network from incomplete data [5, 16]. While it is procedurally very similar to Expectation Maximization (EM) [7, 11], EDML was shown to have certain advantages, both theoretically and practically. Theoretically, EDML can in certain speciali... | [
{
"start": 0,
"end": 1596,
"label": 0
},
{
"start": 1596,
"end": 1780,
"label": 1
},
{
"start": 1780,
"end": 2747,
"label": 0
}
] | 1,596 | 1,780 | Here, we derive, in particular, a new parameter estimation algorithm for Markov networks, which is in many ways a more challenging task, compared to the case of Bayesian networks. | Here, in particular, we derive a new parameter estimation algorithm for Markov networks, a task that is in many ways more challenging than the corresponding case for Bayesian networks. | mixed | 1 | 8fb75ed90f587a43fcfc65dcbe3f73a16c49bbd11eb6826444d208433ef57352 | NeurIPS | 2,013 | https://proceedings.neurips.cc/paper_files/paper/2013/file/b056eb1587586b71e2da9acfe4fbd19e-Paper.pdf | 58c9635f267e65249c13b8c197a0ac80f623ea13fb0c82a47f0e6d2967c2ae8d | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Refaat, Khaled S.",
"Choi, Arthur",
"Darwiche, Adnan"
] | EDML for Learning Parameters in Directed and Undirected Graphical Models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 3}, {"tag": "delete", "original_start": 3, "original_end": 14, "candidate_start": 3, "candidate_end": 3}, {"tag": "equal", "orig... |
b056eb1587586b71e2da9acfe4fbd19e/g01/sentence/paragraph | b056eb1587586b71e2da9acfe4fbd19e | null | b056eb1587586b71e2da9acfe4fbd19e/g01 | authors-3952694096d86f7a | paper_workflow_reconstruction | sentence | paragraph | In this paper, we propose a new perspective on EDML, which views it more abstractly in terms of a simple method for continuous optimization. This new perspective has a number of advantages. First, it makes immediate some results that were previously obtained for EDML, but through some effort. Second, it facilitates the... | [
{
"start": 0,
"end": 468,
"label": 0
},
{
"start": 468,
"end": 652,
"label": 1
},
{
"start": 652,
"end": 870,
"label": 0
}
] | 468 | 652 | Here, we derive, in particular, a new parameter estimation algorithm for Markov networks, which is in many ways a more challenging task, compared to the case of Bayesian networks. | Here, in particular, we derive a new parameter estimation algorithm for Markov networks, a task that is in many ways more challenging than the corresponding case for Bayesian networks. | mixed | 1 | 0f606cf4c1f8724e6a601af7518be43ca6ad80d1d4729d4032692873a28aa4fe | NeurIPS | 2,013 | https://proceedings.neurips.cc/paper_files/paper/2013/file/b056eb1587586b71e2da9acfe4fbd19e-Paper.pdf | 58c9635f267e65249c13b8c197a0ac80f623ea13fb0c82a47f0e6d2967c2ae8d | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Refaat, Khaled S.",
"Choi, Arthur",
"Darwiche, Adnan"
] | EDML for Learning Parameters in Directed and Undirected Graphical Models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 3}, {"tag": "delete", "original_start": 3, "original_end": 14, "candidate_start": 3, "candidate_end": 3}, {"tag": "equal", "orig... |
icml-2013-39ea554c4fd3d608/g01/sentence/context | icml-2013-39ea554c4fd3d608 | null | icml-2013-39ea554c4fd3d608/g01 | authors-3a2a36bbafde89bf | paper_workflow_reconstruction | sentence | context | Bayesian inverse reinforcement learning (Ramachandran & Amir, 2007). In (Dimitrakakis & Rothkopf, 2011) multi-task priors are defined on the policy and the rewards of finite MDPs to learn a different reward function per trajectory without clustering them. A similar approach considers a prior directly on the policy (Dos... | [
{
"start": 0,
"end": 1478,
"label": 0
},
{
"start": 1478,
"end": 1692,
"label": 1
},
{
"start": 1692,
"end": 2686,
"label": 0
}
] | 1,478 | 1,692 | In contrast with the first type of methods, the difficulty here is the segmentation of the time series into an unknown number of local models rather than finding an unknown number of global ones. | For the methods discussed here, the difficulty lies in segmenting the time series into an unknown number of local models, whereas for the first type of methods it lies in finding an unknown number of global models. | mixed | 1 | b1b167fa2a9a50e6734bcbfd7bb49cc97a983a94654f787dc597db9722f08670 | ICML | 2,013 | https://proceedings.mlr.press/v28/almingol13.pdf | dc9c3142315f44f24116bc1aa6863a11f2db1e8ba5d2fb9d0d32bb4ebf2af0cc | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Javier Almingol",
"Lui Montesano",
"Manuel Lopes"
] | Learning Multiple Behaviors from Unlabeled Demonstrations in a Latent Controller Space | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 4, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 4, "original_end": 5, "candidate_start": 1, "candidate_end": 2}, {"tag": "delete", "ori... |
icml-2013-39ea554c4fd3d608/g01/sentence/paragraph | icml-2013-39ea554c4fd3d608 | null | icml-2013-39ea554c4fd3d608/g01 | authors-3a2a36bbafde89bf | paper_workflow_reconstruction | sentence | paragraph | Another different but complementary perspective models each time-series using a set of simpler local models. Several authors approached this problem as a pure regression problem adding local models as needed (Schaal & Atkeson, 1998) or relying on non-parametric methods (Hannah et al., 2011; Rasmussen & Ghahramani, 2002... | [
{
"start": 0,
"end": 486,
"label": 0
},
{
"start": 486,
"end": 700,
"label": 1
},
{
"start": 700,
"end": 799,
"label": 0
}
] | 486 | 700 | In contrast with the first type of methods, the difficulty here is the segmentation of the time series into an unknown number of local models rather than finding an unknown number of global ones. | For the methods discussed here, the difficulty lies in segmenting the time series into an unknown number of local models, whereas for the first type of methods it lies in finding an unknown number of global models. | mixed | 1 | 578032af34fcfa585ba4bd27dfea32077081fb725ab166088ef19b29a0d8cd52 | ICML | 2,013 | https://proceedings.mlr.press/v28/almingol13.pdf | dc9c3142315f44f24116bc1aa6863a11f2db1e8ba5d2fb9d0d32bb4ebf2af0cc | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Javier Almingol",
"Lui Montesano",
"Manuel Lopes"
] | Learning Multiple Behaviors from Unlabeled Demonstrations in a Latent Controller Space | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 4, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 4, "original_end": 5, "candidate_start": 1, "candidate_end": 2}, {"tag": "delete", "ori... |
28dd2c7955ce926456240b2ff0100bde/g01/sentence/context | 28dd2c7955ce926456240b2ff0100bde | null | 28dd2c7955ce926456240b2ff0100bde/g01 | authors-4363a8a344f4f8dc | paper_workflow_reconstruction | sentence | context | We make regret bound comparisons ignoring log log T factors. For generic arm sets, our dependence on d is optimal for linear rewards [34]. For the Bernoulli GLM, our regret has the same order as Zhang et al. [41]. One can show that the regret of Filippi et al. [15] has the same order as ours if we use their assumption ... | [
{
"start": 0,
"end": 1506,
"label": 0
},
{
"start": 1506,
"end": 1664,
"label": 1
},
{
"start": 1664,
"end": 2222,
"label": 0
}
] | 1,506 | 1,664 | GLOC is slightly different from these studies and rather close to Abbasi-Yadkori et al. [2] in that the confidence set is a function of a known regret bound. | GLOC is slightly different from these studies, but rather close to Abbasi-Yadkori et al. [2] in that its confidence set is a function of a known regret bound. | mixed | 1 | 27087a83c91662b6b391a2d6332fb2d9d7015ea8016281a2e074e07b8f0f2cfd | NeurIPS | 2,015 | https://proceedings.neurips.cc/paper_files/paper/2017/file/28dd2c7955ce926456240b2ff0100bde-Paper.pdf | cbaef3a7af11f99efd60e7168feca38ace3aa462454f418fb7870a6cf3d10e40 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Jun, Kwang-Sung",
"Bhargava, Aniruddha",
"Nowak, Robert",
"Willett, Rebecca"
] | Scalable Generalized Linear Bandits: Online Computation and Hashing | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 45, "candidate_start": 0, "candidate_end": 45}, {"tag": "insert", "original_start": 45, "original_end": 45, "candidate_start": 45, "candidate_end": 46}, {"tag": "equal", ... |
28dd2c7955ce926456240b2ff0100bde/g01/sentence/paragraph | 28dd2c7955ce926456240b2ff0100bde | null | 28dd2c7955ce926456240b2ff0100bde/g01 | authors-4363a8a344f4f8dc | paper_workflow_reconstruction | sentence | paragraph | Discussion The trick of obtaining a confidence set from an online learner appeared first in [13, 14] for the linear model, and then was used in [10, 16, 41]. GLOC is slightly different from these studies, but rather close to Abbasi-Yadkori et al. [2] in that its confidence set is a function of a known regret bound. Thi... | [
{
"start": 0,
"end": 158,
"label": 0
},
{
"start": 158,
"end": 316,
"label": 1
},
{
"start": 316,
"end": 488,
"label": 0
}
] | 158 | 316 | GLOC is slightly different from these studies and rather close to Abbasi-Yadkori et al. [2] in that the confidence set is a function of a known regret bound. | GLOC is slightly different from these studies, but rather close to Abbasi-Yadkori et al. [2] in that its confidence set is a function of a known regret bound. | mixed | 1 | c81e2fce0973dd729a3f456e88a4db33c4155567fa47d19f029044e87447caa7 | NeurIPS | 2,015 | https://proceedings.neurips.cc/paper_files/paper/2017/file/28dd2c7955ce926456240b2ff0100bde-Paper.pdf | cbaef3a7af11f99efd60e7168feca38ace3aa462454f418fb7870a6cf3d10e40 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Jun, Kwang-Sung",
"Bhargava, Aniruddha",
"Nowak, Robert",
"Willett, Rebecca"
] | Scalable Generalized Linear Bandits: Online Computation and Hashing | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 45, "candidate_start": 0, "candidate_end": 45}, {"tag": "insert", "original_start": 45, "original_end": 45, "candidate_start": 45, "candidate_end": 46}, {"tag": "equal", ... |
f2e84d98d6dc0c7acd56b40509355666/g01/sentence/context | f2e84d98d6dc0c7acd56b40509355666 | null | f2e84d98d6dc0c7acd56b40509355666/g01 | authors-d8d23842ce4ebf04 | paper_workflow_reconstruction | sentence | context | Note that (A1) and (A2) are very natural. In (A2), if p(i) = 0 for some i ∈ X and q(i) > 0 for some q ∈ Q, then the adversary will never pick q, as the defender can easily detect the presence of the attacker by looking for element i. On the other hand, if p(i) = 0 and q(i) = 0 for all q ∈ Q, we may consider a new alpha... | [
{
"start": 0,
"end": 1052,
"label": 0
},
{
"start": 1052,
"end": 1536,
"label": 1
},
{
"start": 1536,
"end": 2709,
"label": 0
}
] | 1,052 | 1,536 | This is a reasonable model in applications such as multimedia forensics (where one tries to determine if an attacker has tampered with an image from signals that can be modeled as random variables following an image-dependent distribution) or biometrics (where again one tries to detect from random signals whether the p... | This model is reasonable for applications such as multimedia forensics, where the task is to determine whether an attacker has tampered with an image using signals modeled as random variables following a distribution dependent on the image, and biometrics, where detection from random signals again involves determining ... | mixed | 1 | 82ac42c10875e822fafc305e83a4433c08fb6c8eda8903279b14dde1b3e919f7 | NeurIPS | 2,019 | https://proceedings.neurips.cc/paper_files/paper/2019/file/f2e84d98d6dc0c7acd56b40509355666-Paper.pdf | 62441b46f4890bd63ee0a8aa57cec20693df1e9f28f2890c7dbd9fa8feaed1b4 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Yasodharan, Sarath",
"Loiseau, Patrick"
] | Nonzero-sum Adversarial Hypothesis Testing Games | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 5, "candidate_start": 0, "candidate_end": 5}, {"tag": "insert", "original_start": 5, "original_end": 5, "candidate_start": 5, "candidate_end": 11}, {"tag": "equal", "orig... |
f2e84d98d6dc0c7acd56b40509355666/g01/sentence/paragraph | f2e84d98d6dc0c7acd56b40509355666 | null | f2e84d98d6dc0c7acd56b40509355666/g01 | authors-d8d23842ce4ebf04 | paper_workflow_reconstruction | sentence | paragraph | Our setting is that of adversarial hypothesis testing, where the attacker chooses a distribution and points are then generated i.i.d. according to it. This model is reasonable for applications such as multimedia forensics, where the task is to determine whether an attacker has tampered with an image using signals model... | [
{
"start": 0,
"end": 151,
"label": 0
},
{
"start": 151,
"end": 635,
"label": 1
},
{
"start": 635,
"end": 1103,
"label": 0
}
] | 151 | 635 | This is a reasonable model in applications such as multimedia forensics (where one tries to determine if an attacker has tampered with an image from signals that can be modeled as random variables following an image-dependent distribution) or biometrics (where again one tries to detect from random signals whether the p... | This model is reasonable for applications such as multimedia forensics, where the task is to determine whether an attacker has tampered with an image using signals modeled as random variables following a distribution dependent on the image, and biometrics, where detection from random signals again involves determining ... | mixed | 1 | a806d0719edbd4c475c8eeefab0a3f6cea098a3b5b6e77abebf169e7aafc539a | NeurIPS | 2,019 | https://proceedings.neurips.cc/paper_files/paper/2019/file/f2e84d98d6dc0c7acd56b40509355666-Paper.pdf | 62441b46f4890bd63ee0a8aa57cec20693df1e9f28f2890c7dbd9fa8feaed1b4 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Yasodharan, Sarath",
"Loiseau, Patrick"
] | Nonzero-sum Adversarial Hypothesis Testing Games | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 5, "candidate_start": 0, "candidate_end": 5}, {"tag": "insert", "original_start": 5, "original_end": 5, "candidate_start": 5, "candidate_end": 11}, {"tag": "equal", "orig... |
acl-2020-931fb6416cf5d5c3/g01/sentence/context | acl-2020-931fb6416cf5d5c3 | null | acl-2020-931fb6416cf5d5c3/g01 | authors-39213f1c0c600c7a | paper_workflow_reconstruction | sentence | context | In Jabri et al. (2016), the authors introduced a model called Relation Networks, which uses multilayer perceptron models to reason over all pairs of local image features extracted from a grid of image regions. Dynamic tree structures have been used in VQA to capture the visual context of image objects (Tang et al., 201... | [
{
"start": 0,
"end": 812,
"label": 0
},
{
"start": 812,
"end": 1048,
"label": 1
},
{
"start": 1048,
"end": 1894,
"label": 0
}
] | 812 | 1,048 | The graph learner model (Norcliffe-Brown et al., 2018) merges a graph representation of the image based on the question with a graph convolutional network, to learn visual features that can represent question specific interactions. | The graph learner model combines a question-conditioned graph representation of the image with a graph convolutional network to learn visual features capable of representing question-specific interactions (Norcliffe-Brown et al., 2018). | mixed | 1 | e9438ccc12052b0984e12ee8dfd3dba13155f61ad7332300dce954ddd80cd1ed | ACL | 2,020 | https://aclanthology.org/2020.acl-main.643.pdf | a05e2e4ceee0cab1f5f5e71686792bd96393f8d86c646053044fae8ebdce03c3 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Mahmoud Khademi"
] | Multimodal Neural Graph Memory Networks for Visual Question Answering | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 24, "candidate_start": 0, "candidate_end": 24}, {"tag": "replace", "original_start": 24, "original_end": 26, "candidate_start": 24, "candidate_end": 25}, {"tag": "equal",... |
acl-2020-931fb6416cf5d5c3/g01/sentence/paragraph | acl-2020-931fb6416cf5d5c3 | null | acl-2020-931fb6416cf5d5c3/g01 | authors-39213f1c0c600c7a | paper_workflow_reconstruction | sentence | paragraph | Recently, a few models are proposed which can learn the interactions between image regions. The graph learner model combines a question-conditioned graph representation of the image with a graph convolutional network to learn visual features capable of representing question-specific interactions (Norcliffe-Brown et al.... | [
{
"start": 0,
"end": 92,
"label": 0
},
{
"start": 92,
"end": 328,
"label": 1
},
{
"start": 328,
"end": 763,
"label": 0
}
] | 92 | 328 | The graph learner model (Norcliffe-Brown et al., 2018) merges a graph representation of the image based on the question with a graph convolutional network, to learn visual features that can represent question specific interactions. | The graph learner model combines a question-conditioned graph representation of the image with a graph convolutional network to learn visual features capable of representing question-specific interactions (Norcliffe-Brown et al., 2018). | mixed | 1 | d51a4384d9078782560dbfa20e3aa13227f2d244cda23e10bd24e6fa1c2517df | ACL | 2,020 | https://aclanthology.org/2020.acl-main.643.pdf | a05e2e4ceee0cab1f5f5e71686792bd96393f8d86c646053044fae8ebdce03c3 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Mahmoud Khademi"
] | Multimodal Neural Graph Memory Networks for Visual Question Answering | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 24, "candidate_start": 0, "candidate_end": 24}, {"tag": "replace", "original_start": 24, "original_end": 26, "candidate_start": 24, "candidate_end": 25}, {"tag": "equal",... |
acl-2017-eeb27cd3251e137f/g01/sentence/context | acl-2017-eeb27cd3251e137f | null | acl-2017-eeb27cd3251e137f/g01 | authors-062c77bbe0eb645f | paper_workflow_reconstruction | sentence | context | In this section we use word2vec to train real skipgram with negative sampling (SGNS) embedding models. By measuring the value of their objective function and comparing it against the optimal one using exact PMI values, we demonstrate how a well-trained model minimizes the difference in Eq. (10). We note that this is an... | [
{
"start": 0,
"end": 2077,
"label": 0
},
{
"start": 2077,
"end": 2208,
"label": 1
},
{
"start": 2208,
"end": 2731,
"label": 0
}
] | 2,077 | 2,208 | We showed that the optimization of skip-gram embeddings with negative sampling finds the best low-dimensional approximation of the JSMI measure. | We showed that skip-gram embeddings trained with negative sampling find the best low-dimensional approximation of the JSMI measure. | mixed | 1 | cb2b026d6483b87bf5d758c6ec03cd24fe4d86a6de5a720664e3e6adc71f4c3d | ACL | 2,017 | https://aclanthology.org/P17-2026.pdf | f703d5178bd2f3d286aede9d3018c563bff5dbb07a92e4ca4cbc08fb9b3135dd | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Oren Melamud",
"Jacob Goldberger"
] | Information-Theory Interpretation of the Skip-Gram Negative-Sampling Objective Function | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 14, "candidate_start": 0, "candidate_end": 14}, {"tag": "delete", "original_start": 14, "original_end": 34, "candidate_start": 14, "candidate_end": 14}, {"tag": "equal", ... |
acl-2017-eeb27cd3251e137f/g01/sentence/paragraph | acl-2017-eeb27cd3251e137f | null | acl-2017-eeb27cd3251e137f/g01 | authors-062c77bbe0eb645f | paper_workflow_reconstruction | sentence | paragraph | In this study, we developed a new correlation measure between random variables, denoted JSMI. This measure is based on the JS divergence and differs from the standard mutual information measure that is based on the KL divergence. We showed that skip-gram embeddings trained with negative sampling find the best low-dimen... | [
{
"start": 0,
"end": 230,
"label": 0
},
{
"start": 230,
"end": 361,
"label": 1
},
{
"start": 361,
"end": 718,
"label": 0
}
] | 230 | 361 | We showed that the optimization of skip-gram embeddings with negative sampling finds the best low-dimensional approximation of the JSMI measure. | We showed that skip-gram embeddings trained with negative sampling find the best low-dimensional approximation of the JSMI measure. | mixed | 1 | 33ab348a0853252aaab33b1c30e1257dd973d1decd4c23fdb8551df505c2d923 | ACL | 2,017 | https://aclanthology.org/P17-2026.pdf | f703d5178bd2f3d286aede9d3018c563bff5dbb07a92e4ca4cbc08fb9b3135dd | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Oren Melamud",
"Jacob Goldberger"
] | Information-Theory Interpretation of the Skip-Gram Negative-Sampling Objective Function | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 14, "candidate_start": 0, "candidate_end": 14}, {"tag": "delete", "original_start": 14, "original_end": 34, "candidate_start": 14, "candidate_end": 14}, {"tag": "equal", ... |
ad3019b856147c17e82a5bead782d2a8/g01/sentence/context | ad3019b856147c17e82a5bead782d2a8 | null | ad3019b856147c17e82a5bead782d2a8/g01 | authors-4c67e420dfa964a2 | paper_workflow_reconstruction | sentence | context | Throughout this paper we use the formalism for describing transformation-invariant hierarchical architectures developed by Poggio et al. (2012). In [1], the authors propose a theory which, they argue, is general enough to explain the strong performance of convolutional architectures across a wide range of tasks (e.g. [... | [
{
"start": 0,
"end": 910,
"label": 0
},
{
"start": 910,
"end": 1125,
"label": 1
},
{
"start": 1125,
"end": 2070,
"label": 0
}
] | 910 | 1,125 | Our use of computer-generated image datasets lets us completely control the transformations appearing in each test, thereby allowing us to measure properties of the representation for each transformation independently. | We use datasets of computer-generated images, giving us complete control over the transformations appearing in each test and enabling us to measure the representation’s properties separately for each transformation. | mixed | 1 | 626ad7b7e712c2680e4c195286f43996aeef5e4a59337f6ea3e9f5e001ad8eb3 | NeurIPS | 2,013 | https://proceedings.neurips.cc/paper_files/paper/2013/file/ad3019b856147c17e82a5bead782d2a8-Paper.pdf | ae189c6dc60cfe41ec5c87669c956d13bdb187bfdff4c258da25c01db8a731eb | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Liao, Qianli",
"Leibo, Joel Z.",
"Poggio, Tomaso"
] | Learning invariant representations and applications to face verification | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 2}, {"tag": "equal", "original_start": 3, "original_end": 7, "candidate_start": 2, "candidate_end": 6}, {"tag": "insert", "ori... |
ad3019b856147c17e82a5bead782d2a8/g01/sentence/paragraph | ad3019b856147c17e82a5bead782d2a8 | null | ad3019b856147c17e82a5bead782d2a8/g01 | authors-4c67e420dfa964a2 | paper_workflow_reconstruction | sentence | paragraph | The present paper has two primary points. First, we provide empirical support for Poggio et al.’s theory of invariance (which we review in section 2) and show how various pooling methods for convolutional networks can all be understood as building invariance since they are all equivalent to special cases of the model w... | [
{
"start": 0,
"end": 416,
"label": 0
},
{
"start": 416,
"end": 631,
"label": 1
},
{
"start": 631,
"end": 893,
"label": 0
}
] | 416 | 631 | Our use of computer-generated image datasets lets us completely control the transformations appearing in each test, thereby allowing us to measure properties of the representation for each transformation independently. | We use datasets of computer-generated images, giving us complete control over the transformations appearing in each test and enabling us to measure the representation’s properties separately for each transformation. | mixed | 1 | 3defbbef2f1ba6169b3bce8bedb8530ffe87ac22630aaed0f0ab1e002833872f | NeurIPS | 2,013 | https://proceedings.neurips.cc/paper_files/paper/2013/file/ad3019b856147c17e82a5bead782d2a8-Paper.pdf | ae189c6dc60cfe41ec5c87669c956d13bdb187bfdff4c258da25c01db8a731eb | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Liao, Qianli",
"Leibo, Joel Z.",
"Poggio, Tomaso"
] | Learning invariant representations and applications to face verification | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 2}, {"tag": "equal", "original_start": 3, "original_end": 7, "candidate_start": 2, "candidate_end": 6}, {"tag": "insert", "ori... |
09a8a8976abcdfdee15128b4cc02f33a/g01/sentence/context | 09a8a8976abcdfdee15128b4cc02f33a | null | 09a8a8976abcdfdee15128b4cc02f33a/g01 | authors-a1a8b3155c555ad9 | paper_workflow_reconstruction | sentence | context | The stochastic multi-armed bandit (MAB) problem is a sequential decision-making problem where, at each step of a T -period run, a learner plays one of k arms and observes a corresponding loss that is sampled independently from an underlying distribution with unknown parameters. The learner’s goal is to minimize the pse... | [
{
"start": 0,
"end": 1348,
"label": 0
},
{
"start": 1348,
"end": 1660,
"label": 1
},
{
"start": 1660,
"end": 2734,
"label": 0
}
] | 1,348 | 1,660 | They are also naturally relevant for many cyber-physical systems with humans in the loop (e.g., pricing end-use demand in societalscale infrastructure systems such as power grids or transportation networks to minimize system costs given the limited number of user interactions possible). | They are also naturally relevant to many cyber-physical systems involving humans in the loop; for example, end-use demand may be priced in societal-scale infrastructure systems, such as power grids and transportation networks, to minimize system costs when only a limited number of user interactions is possible. | mixed | 1 | b49edc7e89044de2a0df2632ad0646ce54789e9687fd6da96d30b1dda31fcd07 | NeurIPS | 2,019 | https://proceedings.neurips.cc/paper_files/paper/2019/file/09a8a8976abcdfdee15128b4cc02f33a-Paper.pdf | 90a33c31e1844cbe8458f72e6ccf9e81a1d3e185c415bc9d7d93cb0a69610ed9 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Amani, Sanae",
"Alizadeh, Mahnoosh",
"Thrampoulidis, Christos"
] | Linear Stochastic Bandits Under Safety Constraints | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 33, "candidate_start": 0, "candidate_end": 33}, {"tag": "replace", "original_start": 33, "original_end": 34, "candidate_start": 33, "candidate_end": 34}, {"tag": "equal",... |
09a8a8976abcdfdee15128b4cc02f33a/g01/sentence/paragraph | 09a8a8976abcdfdee15128b4cc02f33a | null | 09a8a8976abcdfdee15128b4cc02f33a/g01 | authors-a1a8b3155c555ad9 | paper_workflow_reconstruction | sentence | paragraph | Bandit algorithms have found many applications in systems that repeatedly deal with unknown stochastic environments (such as humans) and seek to optimize a long-term reward by simultaneously learning and exploiting the unknown environment (e.g., ad display optimization algorithms with unknown user preferences, path rou... | [
{
"start": 0,
"end": 354,
"label": 0
},
{
"start": 354,
"end": 666,
"label": 1
},
{
"start": 666,
"end": 1240,
"label": 0
}
] | 354 | 666 | They are also naturally relevant for many cyber-physical systems with humans in the loop (e.g., pricing end-use demand in societalscale infrastructure systems such as power grids or transportation networks to minimize system costs given the limited number of user interactions possible). | They are also naturally relevant to many cyber-physical systems involving humans in the loop; for example, end-use demand may be priced in societal-scale infrastructure systems, such as power grids and transportation networks, to minimize system costs when only a limited number of user interactions is possible. | mixed | 1 | 98729596ebde73456c4c71c001d4897208e5b66fa37aafcfbcc7ccbedde62688 | NeurIPS | 2,019 | https://proceedings.neurips.cc/paper_files/paper/2019/file/09a8a8976abcdfdee15128b4cc02f33a-Paper.pdf | 90a33c31e1844cbe8458f72e6ccf9e81a1d3e185c415bc9d7d93cb0a69610ed9 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Amani, Sanae",
"Alizadeh, Mahnoosh",
"Thrampoulidis, Christos"
] | Linear Stochastic Bandits Under Safety Constraints | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 33, "candidate_start": 0, "candidate_end": 33}, {"tag": "replace", "original_start": 33, "original_end": 34, "candidate_start": 33, "candidate_end": 34}, {"tag": "equal",... |
acl-2015-97ceef6824205c8f/g01/sentence/context | acl-2015-97ceef6824205c8f | null | acl-2015-97ceef6824205c8f/g01 | authors-7094f7e384bc1479 | paper_workflow_reconstruction | sentence | context | Any summarisation evaluation based on human judgment is inherently subjective, but we restrict this subjectivity in three ways. First, timeline creation (step 1) involves the selection of important content, which is by far the most subjective of the decisions involved in our evaluation method. We therefore ask three in... | [
{
"start": 0,
"end": 2283,
"label": 0
},
{
"start": 2283,
"end": 2385,
"label": 1
},
{
"start": 2385,
"end": 3461,
"label": 0
}
] | 2,283 | 2,385 | For system evaluation, this means that a system can score at most one point per HCU (exactly one point if it gives a perfect rendition of that HCU). | For system evaluation, each HCU permits a maximum score of one point, awarded for a perfect rendition. | mixed | 1 | 4dabf82fb0afaffdca7f703b48f6d269d5ab96144890423ff75304ea82ae0a45 | ACL | 2,015 | https://aclanthology.org/P15-2137.pdf | 1c67ee2143692360d7af1888807b13b99642c9582acdc132e9df47cce2b2baf2 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Sandro Bauer",
"Simone Teufel"
] | A Methodology for Evaluating Timeline Generation Algorithms based on Deep Semantic Units | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 23, "candidate_start": 0, "candidate_end": 23}, {"tag": "delete", "original_start": 23, "original_end": 29, "candidate_start": 23, "candidate_end": 23}, {"tag": "equal", ... |
acl-2015-97ceef6824205c8f/g01/sentence/paragraph | acl-2015-97ceef6824205c8f | null | acl-2015-97ceef6824205c8f/g01 | authors-7094f7e384bc1479 | paper_workflow_reconstruction | sentence | paragraph | HCUs are more abstract in that they express an event, i.e. a concrete real-world action (France invades Algeria) or state change (Obama becomes president), while SCUs are more textual, not semantically defined and generally represent a smaller unit of meaning. State descriptions, opinions, wishes, aspirations, intentio... | [
{
"start": 0,
"end": 851,
"label": 0
},
{
"start": 851,
"end": 953,
"label": 1
},
{
"start": 953,
"end": 1191,
"label": 0
}
] | 851 | 953 | For system evaluation, this means that a system can score at most one point per HCU (exactly one point if it gives a perfect rendition of that HCU). | For system evaluation, each HCU permits a maximum score of one point, awarded for a perfect rendition. | mixed | 1 | 5d649448c0a134fd37aec58d8f54097becb820ae50bc0e53e594844b219bc7d7 | ACL | 2,015 | https://aclanthology.org/P15-2137.pdf | 1c67ee2143692360d7af1888807b13b99642c9582acdc132e9df47cce2b2baf2 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Sandro Bauer",
"Simone Teufel"
] | A Methodology for Evaluating Timeline Generation Algorithms based on Deep Semantic Units | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 23, "candidate_start": 0, "candidate_end": 23}, {"tag": "delete", "original_start": 23, "original_end": 29, "candidate_start": 23, "candidate_end": 23}, {"tag": "equal", ... |
acl-2014-77361daea062f65e/g01/sentence/context | acl-2014-77361daea062f65e | null | acl-2014-77361daea062f65e/g01 | authors-897f6024bac663cc | paper_workflow_reconstruction | sentence | context | As stated earlier, our brute-force baseline approach is to parse the sentence using a full contextfree grammar (CFG) and then hedge-transform the result. This method should yield a ceiling on hedge-parsing accuracy, as it has access to rich contextual information (as compared to grammars trained on transformed trees). ... | [
{
"start": 0,
"end": 944,
"label": 0
},
{
"start": 944,
"end": 1099,
"label": 1
},
{
"start": 1099,
"end": 2124,
"label": 0
}
] | 944 | 1,099 | Other than the symbol at the root of the tree, the only constituents with span length greater than L in the binarized tree will be labeled with these special binarization non-terminals. | In the binarized tree, constituents of span length greater than L will be the only ones assigned binarization non-terminals, with the root symbol excepted. | mixed | 1 | 2e198beb8d549ca64ba0364cf26a6f75a905e20c92ec0d98a3b39e0090a33258 | ACL | 2,014 | https://aclanthology.org/P14-2129.pdf | aa0a6228925ce2efe2f4ad98a2f3c94a05ee196b96109c3e77f1ebf8fe0ac161 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Mahsa Yarmohammadi",
"Aaron Dunlop",
"Brian Roark"
] | Transforming trees into hedges and parsing with “hedgebank” grammars | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 9, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 9, "original_end": 15, "candidate_start": 1, "candidate_end": 7}, {"tag": "delete", "or... |
acl-2014-77361daea062f65e/g01/sentence/paragraph | acl-2014-77361daea062f65e | null | acl-2014-77361daea062f65e/g01 | authors-897f6024bac663cc | paper_workflow_reconstruction | sentence | paragraph | Since we limit the span of non-terminal labels, we can constrain the search performed by the parser, greatly reduce the CYK processing time. In essence, we perform no work in chart cells spanning more than L words, except for the cells along the periphery of the chart, which are just used to connect the hedges to the r... | [
{
"start": 0,
"end": 508,
"label": 0
},
{
"start": 508,
"end": 663,
"label": 1
},
{
"start": 663,
"end": 1017,
"label": 0
}
] | 508 | 663 | Other than the symbol at the root of the tree, the only constituents with span length greater than L in the binarized tree will be labeled with these special binarization non-terminals. | In the binarized tree, constituents of span length greater than L will be the only ones assigned binarization non-terminals, with the root symbol excepted. | mixed | 1 | 8fe1975a89aaabcde74760b21140af91186dd4a9ed663eca927818f6b861bb12 | ACL | 2,014 | https://aclanthology.org/P14-2129.pdf | aa0a6228925ce2efe2f4ad98a2f3c94a05ee196b96109c3e77f1ebf8fe0ac161 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Mahsa Yarmohammadi",
"Aaron Dunlop",
"Brian Roark"
] | Transforming trees into hedges and parsing with “hedgebank” grammars | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 9, "candidate_start": 0, "candidate_end": 1}, {"tag": "equal", "original_start": 9, "original_end": 15, "candidate_start": 1, "candidate_end": 7}, {"tag": "delete", "or... |
cdfa4c42f465a5a66871587c69fcfa34/g01/sentence/context | cdfa4c42f465a5a66871587c69fcfa34 | null | cdfa4c42f465a5a66871587c69fcfa34/g01 | authors-fc749d83b723e902 | paper_workflow_reconstruction | sentence | context | We also run experiments to verify the performance of our proposed algorithms in practical machine learning applications. We observe that our algorithms match the performance of the prior work, while providing either a better guaranteed approximation or a better running time.
Further Related Work. Buchbinder and Feldma... | [
{
"start": 0,
"end": 866,
"label": 0
},
{
"start": 866,
"end": 983,
"label": 1
},
{
"start": 983,
"end": 1819,
"label": 0
}
] | 866 | 983 | Chen et al. [2018], Roughgarden and Wang [2018] consider submodular maximization in the context of online no-regret learning. | Chen et al. [2018] and Roughgarden and Wang [2018] consider submodular maximization within online no-regret learning. | mixed | 1 | 377651fe4811e53f4db2a224b47ccc8cea6311bb7c5251dbcbbb6a6a4db59789 | NeurIPS | 2,018 | https://proceedings.neurips.cc/paper_files/paper/2018/file/cdfa4c42f465a5a66871587c69fcfa34-Paper.pdf | 881b793316732747ae60ff631cca923442ab66ba3ee3cec9f3b35e5bd5015c5d | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Niazadeh, Rad",
"Roughgarden, Tim",
"Wang, Joshua"
] | Optimal Algorithms for Continuous Non-monotone Submodular and DR-Submodular Maximization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 18, "candidate_start": 0, "candidate_end": 18}, {"tag": "replace", "original_start": 18, "original_end": 19, "candidate_start": 18, "candidate_end": 22}, {"tag": "equal",... |
cdfa4c42f465a5a66871587c69fcfa34/g01/sentence/paragraph | cdfa4c42f465a5a66871587c69fcfa34 | null | cdfa4c42f465a5a66871587c69fcfa34/g01 | authors-fc749d83b723e902 | paper_workflow_reconstruction | sentence | paragraph | Further Related Work. Buchbinder and Feldman [2016] derandomize the bi-greedy algorithm. Staib and Jegelka [2017] apply continuous submodular optimization to budget allocation, and develop a new submodular optimization algorithm to this end. Hassani et al. [2017] give a 12 -approximation for monotone continuous DR-subm... | [
{
"start": 0,
"end": 589,
"label": 0
},
{
"start": 589,
"end": 706,
"label": 1
},
{
"start": 706,
"end": 963,
"label": 0
}
] | 589 | 706 | Chen et al. [2018], Roughgarden and Wang [2018] consider submodular maximization in the context of online no-regret learning. | Chen et al. [2018] and Roughgarden and Wang [2018] consider submodular maximization within online no-regret learning. | mixed | 1 | e5d4de0410518c3d7bf1bd7d307417bc08bf8075684a73280fe46434a370fdf7 | NeurIPS | 2,018 | https://proceedings.neurips.cc/paper_files/paper/2018/file/cdfa4c42f465a5a66871587c69fcfa34-Paper.pdf | 881b793316732747ae60ff631cca923442ab66ba3ee3cec9f3b35e5bd5015c5d | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Niazadeh, Rad",
"Roughgarden, Tim",
"Wang, Joshua"
] | Optimal Algorithms for Continuous Non-monotone Submodular and DR-Submodular Maximization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 18, "candidate_start": 0, "candidate_end": 18}, {"tag": "replace", "original_start": 18, "original_end": 19, "candidate_start": 18, "candidate_end": 22}, {"tag": "equal",... |
icml-2017-7aa86556beca6029/g01/sentence/context | icml-2017-7aa86556beca6029 | null | icml-2017-7aa86556beca6029/g01 | authors-afd6637e3e5f4c8c | paper_workflow_reconstruction | sentence | context | In this section we describe simulation experiments to evaluate the running time of GDUAL - FORWARD against other algorithms, and to assess the ability to learn a wide variety of models for which exact likelihood calculations were not previously possible, by using GDUAL - FORWARD within a parameter estimation routine.
... | [
{
"start": 0,
"end": 535,
"label": 0
},
{
"start": 535,
"end": 722,
"label": 1
},
{
"start": 722,
"end": 1913,
"label": 0
}
] | 535 | 722 | PGF - FORWARD is only applicable to the Poisson HMM from (Winner & Sheldon, 2016), which, in our terminology, is a model with a Poisson immigration distribution and a Bernoulli offspring distribution. | PGF-FORWARD applies only to the Poisson HMM of Winner & Sheldon (2016); following our terminology, the model has a Poisson immigration distribution and a Bernoulli offspring distribution. | mixed | 1 | 6e07032fb35690a1b7afb140eea2ff2332a786f0ed9c3923b3340c159bfd45cf | ICML | 2,017 | https://proceedings.mlr.press/v70/winner17a/winner17a.pdf | ce6d418ca9836ffe4f2db111d93a4446621f8812cc77e85223b104c6ec3b0b98 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Kevin Winner",
"Debora Sujono",
"Dan Sheldon"
] | Exact Inference for Integer Latent-Variable Models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 3}, {"tag": "delete", "original_start": 3, "original_end": 4, "candidate_start": 3, "candidate_end": 3}, {"tag": "equal", "origi... |
icml-2017-7aa86556beca6029/g01/sentence/paragraph | icml-2017-7aa86556beca6029 | null | icml-2017-7aa86556beca6029/g01 | authors-afd6637e3e5f4c8c | paper_workflow_reconstruction | sentence | paragraph | Running time vs Y . We compared the running time of GDUAL - FORWARD with the PGF - FORWARD algorithm from (Winner & Sheldon, 2016) as well as T RUNC , the standard truncated forward algorithm (Dail & Madsen, 2011). PGF-FORWARD applies only to the Poisson HMM of Winner & Sheldon (2016); following our terminology, the mo... | [
{
"start": 0,
"end": 215,
"label": 0
},
{
"start": 215,
"end": 402,
"label": 1
},
{
"start": 402,
"end": 594,
"label": 0
}
] | 215 | 402 | PGF - FORWARD is only applicable to the Poisson HMM from (Winner & Sheldon, 2016), which, in our terminology, is a model with a Poisson immigration distribution and a Bernoulli offspring distribution. | PGF-FORWARD applies only to the Poisson HMM of Winner & Sheldon (2016); following our terminology, the model has a Poisson immigration distribution and a Bernoulli offspring distribution. | mixed | 1 | 5adf8b3de12764675c2a2fabef448469ec15cd92bc5d9b6133b14ac199afa54c | ICML | 2,017 | https://proceedings.mlr.press/v70/winner17a/winner17a.pdf | ce6d418ca9836ffe4f2db111d93a4446621f8812cc77e85223b104c6ec3b0b98 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Kevin Winner",
"Debora Sujono",
"Dan Sheldon"
] | Exact Inference for Integer Latent-Variable Models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 3, "candidate_start": 0, "candidate_end": 3}, {"tag": "delete", "original_start": 3, "original_end": 4, "candidate_start": 3, "candidate_end": 3}, {"tag": "equal", "origi... |
icml-2013-34a4bbb8fc0f9faf/g01/sentence/context | icml-2013-34a4bbb8fc0f9faf | null | icml-2013-34a4bbb8fc0f9faf/g01 | authors-5c6d8ce933ab4fe1 | paper_workflow_reconstruction | sentence | context | The value kE n k Ξ is of obvious importance. E n is the minimiser of the empirical regularised risk, and if this minimiser converges with high probability to the minimiser of the regularised risk, then one can infer from Caponnetto De Vito (2007)[Prop. 3] that E n will be bounded with high probability. This then guaran... | [
{
"start": 0,
"end": 594,
"label": 0
},
{
"start": 594,
"end": 705,
"label": 1
},
{
"start": 705,
"end": 1309,
"label": 0
}
] | 594 | 705 | Derivations of estimates are often generic, and operations can naturally be combined to form complex estimates. | Derivations of estimates are often generic, and operations can be combined naturally to form complex estimates. | mixed | 1 | cc63d9127aa57b3d94939850a762dab9c377118062cd4ca8fa43d5bd461a1860 | ICML | 2,013 | https://proceedings.mlr.press/v28/grunewalder13.pdf | b383e24ce5d8baae302d8b7cefae824b401721356c78169d8b548683fb17fc22 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Steffen Grunewalder",
"Gretton Arthur",
"John Shawe-Taylor"
] | Smooth Operators | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 63, "candidate_start": 0, "candidate_end": 63}, {"tag": "delete", "original_start": 63, "original_end": 73, "candidate_start": 63, "candidate_end": 63}, {"tag": "equal", ... |
icml-2013-34a4bbb8fc0f9faf/g01/sentence/paragraph | icml-2013-34a4bbb8fc0f9faf | null | icml-2013-34a4bbb8fc0f9faf/g01 | authors-5c6d8ce933ab4fe1 | paper_workflow_reconstruction | sentence | paragraph | We have presented an approach for estimating linear operators acting on an RKHS. Derivations of estimates are often generic, and operations can be combined naturally to form complex estimates. Risk bounds for these complex rules can be expressed straightforwardly in terms of risk bounds of the basic estimates used in b... | [
{
"start": 0,
"end": 81,
"label": 0
},
{
"start": 81,
"end": 192,
"label": 1
},
{
"start": 192,
"end": 597,
"label": 0
}
] | 81 | 192 | Derivations of estimates are often generic, and operations can naturally be combined to form complex estimates. | Derivations of estimates are often generic, and operations can be combined naturally to form complex estimates. | mixed | 1 | c846e2be26afac9dda7cc8d5693bdb634d2d6fac10571dd63d0bdf546a9c508d | ICML | 2,013 | https://proceedings.mlr.press/v28/grunewalder13.pdf | b383e24ce5d8baae302d8b7cefae824b401721356c78169d8b548683fb17fc22 | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Steffen Grunewalder",
"Gretton Arthur",
"John Shawe-Taylor"
] | Smooth Operators | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 63, "candidate_start": 0, "candidate_end": 63}, {"tag": "delete", "original_start": 63, "original_end": 73, "candidate_start": 63, "candidate_end": 63}, {"tag": "equal", ... |
acl-2016-1dbf1028b8496700/g01/sentence/context | acl-2016-1dbf1028b8496700 | null | acl-2016-1dbf1028b8496700/g01 | authors-a5ad4495e059f4d1 | paper_workflow_reconstruction | sentence | context | Socher et al. (2013), Kalchbrenner et al. (2014) and Kim (2014)). These representations are automatically trained from data and can simultaneously encode multiple linguistics features. For example, word embeddings can encode gender distinctions and plural-singular distinctions (Mikolov et al., 2013b) and the representa... | [
{
"start": 0,
"end": 2985,
"label": 0
},
{
"start": 2985,
"end": 3130,
"label": 1
},
{
"start": 3130,
"end": 4145,
"label": 0
}
] | 2,985 | 3,130 | The inclusion of syntactic structures in the feature set provided a boost to the performance of the model trained on global lexical context and discourse cohesion. | Adding syntactic structures to the feature set improved the performance of the model trained using global lexical context and discourse cohesion. | mixed | 1 | 5cbc0d28ef377bf5a9ddf60395dd66701dc96826d83cbf33f5f94afa0ecf2550 | ACL | 2,016 | https://aclanthology.org/P16-1019.pdf | e8aec5b7cd43a4cb27771bd7d0ec4f7954903b2feacf504617fdafa3ebd908da | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Giancarlo Salton",
"Robert Ross",
"John Kelleher"
] | Idiom Token Classification using Sentential Distributed Semantics | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 4, "candidate_start": 0, "candidate_end": 3}, {"tag": "equal", "original_start": 4, "original_end": 6, "candidate_start": 3, "candidate_end": 5}, {"tag": "replace", "or... |
acl-2016-1dbf1028b8496700/g01/sentence/paragraph | acl-2016-1dbf1028b8496700 | null | acl-2016-1dbf1028b8496700/g01 | authors-a5ad4495e059f4d1 | paper_workflow_reconstruction | sentence | paragraph | Focusing on idiom token classification in English, Fazly et al. (2009) developed the concept of a canonical form (defined in terms of local syntactic and lexical patterns) and argued that for each idiom there is a distinct canonical form (or small set of forms) that mark idiomatic usages of a phrase. Meanwhile Sporlede... | [
{
"start": 0,
"end": 1234,
"label": 0
},
{
"start": 1234,
"end": 1379,
"label": 1
},
{
"start": 1379,
"end": 1686,
"label": 0
}
] | 1,234 | 1,379 | The inclusion of syntactic structures in the feature set provided a boost to the performance of the model trained on global lexical context and discourse cohesion. | Adding syntactic structures to the feature set improved the performance of the model trained using global lexical context and discourse cohesion. | mixed | 1 | bf5030581a93247872a21e17d6eea0fa891170deefe160bed8a399c73182b333 | ACL | 2,016 | https://aclanthology.org/P16-1019.pdf | e8aec5b7cd43a4cb27771bd7d0ec4f7954903b2feacf504617fdafa3ebd908da | null | null | known_replacement_provenance | false | 1 | 1 | true | openai/gpt-6-luna | validation | [
"Giancarlo Salton",
"Robert Ross",
"John Kelleher"
] | Idiom Token Classification using Sentential Distributed Semantics | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 4, "candidate_start": 0, "candidate_end": 3}, {"tag": "equal", "original_start": 4, "original_end": 6, "candidate_start": 3, "candidate_end": 5}, {"tag": "replace", "or... |
acl-2015-653bd4ed07f4a30a/g01/sentence/context | acl-2015-653bd4ed07f4a30a | null | acl-2015-653bd4ed07f4a30a/g01 | authors-8d12bc9e9f917c74 | paper_workflow_reconstruction | sentence | context | Semi-supervised experiments Because our PYHSMM is a generative model, it is easily amenable to semi-supervised segmentation and tagging. We used random 10,000 sentences from supervised data on Kyoto, BCCWJ, and LWC datasets along with unsupervised datasets in Table 3. Results are shown in Table 6: segmentation accuraci... | [
{
"start": 0,
"end": 1166,
"label": 0
},
{
"start": 1166,
"end": 1371,
"label": 1
},
{
"start": 1371,
"end": 2181,
"label": 0
}
] | 1,166 | 1,371 | Even from the surface appearance of characters, we can see that similar words are assigned to the same state including some emoticons (states 9,29,32), and in fact we can identify a state of postpositions specific to that dialect (state 3). | Based solely on character-surface appearance, we see similar words, including some emoticons, grouped in states 9, 29, and 32; in fact, we identify state 3 as a postposition state specific to this dialect. | mixed | 1 | 210f42768c4e10980cc9b07c8e3168e91d1d93ccc88831a36e82762a99e78bcd | ACL | 2,015 | https://aclanthology.org/P15-1171.pdf | 9a855cde232ad9966ac9462dfb7883ed5393e8b1c59bb92596e59a989dbeb2ad | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Kei Uchiumi",
"Hiroshi Tsukahara",
"Daichi Mochihashi"
] | Inducing Word and Part-of-Speech with Pitman-Yor Hidden Semi-Markov Models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 2, "candidate_start": 0, "candidate_end": 3}, {"tag": "equal", "original_start": 2, "original_end": 3, "candidate_start": 3, "candidate_end": 4}, {"tag": "insert", "ori... |
acl-2015-653bd4ed07f4a30a/g01/sentence/paragraph | acl-2015-653bd4ed07f4a30a | null | acl-2015-653bd4ed07f4a30a/g01 | authors-8d12bc9e9f917c74 | paper_workflow_reconstruction | sentence | paragraph | Finally, we note that our proposed model for unsupervised learning is most effective for the language which we do not know its syntactic behavior but only know raw strings as its data. In Figure 8, we show an excerpt of results to model a Japanese local dialect (Mikawa-ben around Nagoya district) collected from a speci... | [
{
"start": 0,
"end": 333,
"label": 0
},
{
"start": 333,
"end": 538,
"label": 1
},
{
"start": 538,
"end": 881,
"label": 0
}
] | 333 | 538 | Even from the surface appearance of characters, we can see that similar words are assigned to the same state including some emoticons (states 9,29,32), and in fact we can identify a state of postpositions specific to that dialect (state 3). | Based solely on character-surface appearance, we see similar words, including some emoticons, grouped in states 9, 29, and 32; in fact, we identify state 3 as a postposition state specific to this dialect. | mixed | 1 | d9bb0e1e9f10b52239f4566afe47401f09ed0bd3693c043462d0cde004382e8c | ACL | 2,015 | https://aclanthology.org/P15-1171.pdf | 9a855cde232ad9966ac9462dfb7883ed5393e8b1c59bb92596e59a989dbeb2ad | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Kei Uchiumi",
"Hiroshi Tsukahara",
"Daichi Mochihashi"
] | Inducing Word and Part-of-Speech with Pitman-Yor Hidden Semi-Markov Models | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "replace", "original_start": 0, "original_end": 2, "candidate_start": 0, "candidate_end": 3}, {"tag": "equal", "original_start": 2, "original_end": 3, "candidate_start": 3, "candidate_end": 4}, {"tag": "insert", "ori... |
97d98119037c5b8a9663cb21fb8ebf47/g01/sentence/context | 97d98119037c5b8a9663cb21fb8ebf47 | null | 97d98119037c5b8a9663cb21fb8ebf47/g01 | authors-915c78c7f93236ce | paper_workflow_reconstruction | sentence | context | Our specific focus is on providing formal guarantees on future executions of control algorithms in terms of maximum expected cost (quantifying performance) and maximum probability of constraint violation (quantifying safety). Such bounds determine the reliability of control in the presence of process, measurement and p... | [
{
"start": 0,
"end": 1242,
"label": 0
},
{
"start": 1242,
"end": 1534,
"label": 1
},
{
"start": 1534,
"end": 2407,
"label": 0
}
] | 1,242 | 1,534 | This is accomplished using concentration-of-measure inequalities [9] which provide only probabilistic bounds , i.e. they certify the algorithm execution in terms of statements such as: “in future executions, with 99% chance the expected cost will be less than X and the probability of collision will be less than Y”. | This is accomplished through concentration-of-measure inequalities [9], which provide only probabilistic bounds that certify algorithm execution—for example, that across future executions there is a 99% chance that expected cost is less than X and the probability of collision is less than Y. | mixed | 1 | 5fb00a880578b0a50ca1efa3775737a96186a6c519b8d6356157ebe345c44ab9 | NeurIPS | 2,015 | https://proceedings.neurips.cc/paper_files/paper/2015/file/97d98119037c5b8a9663cb21fb8ebf47-Paper.pdf | ab3a1df80f68615fc3705df9d0cf9efa9472a8a2b4b47f8cb7d68fcd4909c4b6 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Kobilarov, Marin"
] | Sample Complexity Bounds for Iterative Stochastic Policy Optimization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 21, "candidate_start": 0, "candidate_end": 21}, {"tag": "insert", "original_start": 21, "original_end": 21, "candidate_start": 21, "candidate_end": 25}, {"tag": "equal", ... |
97d98119037c5b8a9663cb21fb8ebf47/g01/sentence/paragraph | 97d98119037c5b8a9663cb21fb8ebf47 | null | 97d98119037c5b8a9663cb21fb8ebf47/g01 | authors-915c78c7f93236ce | paper_workflow_reconstruction | sentence | paragraph | Approach. To cope with such general conditions, we study robustness through a statistical learning viewpoint [6, 7, 8] using finite-time sample complexity bounds on performance based on empirical runs. This is accomplished through concentration-of-measure inequalities [9], which provide only probabilistic bounds that c... | [
{
"start": 0,
"end": 202,
"label": 0
},
{
"start": 202,
"end": 494,
"label": 1
},
{
"start": 494,
"end": 648,
"label": 0
}
] | 202 | 494 | This is accomplished using concentration-of-measure inequalities [9] which provide only probabilistic bounds , i.e. they certify the algorithm execution in terms of statements such as: “in future executions, with 99% chance the expected cost will be less than X and the probability of collision will be less than Y”. | This is accomplished through concentration-of-measure inequalities [9], which provide only probabilistic bounds that certify algorithm execution—for example, that across future executions there is a 99% chance that expected cost is less than X and the probability of collision is less than Y. | mixed | 1 | 589c11580bf47a6efbf4a148cb81789c5ef410fa47982b9fac788424035382eb | NeurIPS | 2,015 | https://proceedings.neurips.cc/paper_files/paper/2015/file/97d98119037c5b8a9663cb21fb8ebf47-Paper.pdf | ab3a1df80f68615fc3705df9d0cf9efa9472a8a2b4b47f8cb7d68fcd4909c4b6 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Kobilarov, Marin"
] | Sample Complexity Bounds for Iterative Stochastic Policy Optimization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 21, "candidate_start": 0, "candidate_end": 21}, {"tag": "insert", "original_start": 21, "original_end": 21, "candidate_start": 21, "candidate_end": 25}, {"tag": "equal", ... |
icml-2019-ad21da0671d9786f/g01/sentence/context | icml-2019-ad21da0671d9786f | null | icml-2019-ad21da0671d9786f/g01 | authors-5cac4e4a2328d9a7 | paper_workflow_reconstruction | sentence | context | On the left, ξ is a random variable representing a data point drawn according to some distribution and f ̃ (x, ξ) measures the fit of some model parameter x to the data point ξ.
Even though this pessimistic result applies to the general stochastic case, linear convergence rates can be obtained for deterministic finite... | [
{
"start": 0,
"end": 327,
"label": 0
},
{
"start": 327,
"end": 698,
"label": 1
},
{
"start": 698,
"end": 1726,
"label": 0
}
] | 327 | 698 | For instance, linear rates are achieved by SAG (Schmidt et al., 2017), SAGA (Defazio et al., 2014), SVRG (Johnson & Zhang, 2013; Xiao & Zhang, 2014), SDCA (Shalev-Shwartz & Zhang, 2016), MISO (Mairal, 2015), Katyusha (Allen-Zhu, 2017), MiG (Zhou et al., 2018), SARAH (Nguyen et al., 2017), accelerated SAGA (Zhou, 2019) ... | For instance, linear rates are achieved by SAG (Schmidt et al., 2017), SAGA (Defazio et al., 2014), SVRG (Johnson & Zhang, 2013; Xiao & Zhang, 2014), SDCA (Shalev-Shwartz & Zhang, 2016), MISO (Mairal, 2015), Katyusha (Allen-Zhu, 2017), MiG (Zhou et al., 2018), SARAH (Nguyen et al., 2017), accelerated SAGA (Zhou, 2019),... | mixed | 1 | 7e3ae695240abc9cf04483117e17c5a058560e41dd16e8684cf8c480b7cab449 | ICML | 2,019 | https://proceedings.mlr.press/v97/kulunchakov19a/kulunchakov19a.pdf | 7385d33497eb621ebbbc92b4b020dbf892745304a6faa8657560c504de031a28 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Andrei Kulunchakov",
"Julien Mairal"
] | Estimate Sequences for Variance-Reduced Stochastic Composite Optimization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 319, "candidate_start": 0, "candidate_end": 319}, {"tag": "insert", "original_start": 319, "original_end": 319, "candidate_start": 319, "candidate_end": 320}, {"tag": "eq... |
icml-2019-ad21da0671d9786f/g01/sentence/paragraph | icml-2019-ad21da0671d9786f | null | icml-2019-ad21da0671d9786f/g01 | authors-5cac4e4a2328d9a7 | paper_workflow_reconstruction | sentence | paragraph | Even though this pessimistic result applies to the general stochastic case, linear convergence rates can be obtained for deterministic finite sums. For instance, linear rates are achieved by SAG (Schmidt et al., 2017), SAGA (Defazio et al., 2014), SVRG (Johnson & Zhang, 2013; Xiao & Zhang, 2014), SDCA (Shalev-Shwartz &... | [
{
"start": 0,
"end": 148,
"label": 0
},
{
"start": 148,
"end": 519,
"label": 1
},
{
"start": 519,
"end": 883,
"label": 0
}
] | 148 | 519 | For instance, linear rates are achieved by SAG (Schmidt et al., 2017), SAGA (Defazio et al., 2014), SVRG (Johnson & Zhang, 2013; Xiao & Zhang, 2014), SDCA (Shalev-Shwartz & Zhang, 2016), MISO (Mairal, 2015), Katyusha (Allen-Zhu, 2017), MiG (Zhou et al., 2018), SARAH (Nguyen et al., 2017), accelerated SAGA (Zhou, 2019) ... | For instance, linear rates are achieved by SAG (Schmidt et al., 2017), SAGA (Defazio et al., 2014), SVRG (Johnson & Zhang, 2013; Xiao & Zhang, 2014), SDCA (Shalev-Shwartz & Zhang, 2016), MISO (Mairal, 2015), Katyusha (Allen-Zhu, 2017), MiG (Zhou et al., 2018), SARAH (Nguyen et al., 2017), accelerated SAGA (Zhou, 2019),... | mixed | 1 | 5121dd9c4333161c40a117a429c4c287a5b67d938b4d0561070bfb958ab026c0 | ICML | 2,019 | https://proceedings.mlr.press/v97/kulunchakov19a/kulunchakov19a.pdf | 7385d33497eb621ebbbc92b4b020dbf892745304a6faa8657560c504de031a28 | null | null | known_replacement_provenance | false | 1 | 1 | false | openai/gpt-6-luna | validation | [
"Andrei Kulunchakov",
"Julien Mairal"
] | Estimate Sequences for Variance-Reduced Stochastic Composite Optimization | {"split": "calibration", "protocol_version": 3, "source_quality_flags": [], "edit_operations": [{"tag": "equal", "original_start": 0, "original_end": 319, "candidate_start": 0, "candidate_end": 319}, {"tag": "insert", "original_start": 319, "original_end": 319, "candidate_start": 319, "candidate_end": 320}, {"tag": "eq... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.