Redraft 8 flagged questions (source-grounded, de-giveaway): q64 q85 q102 q104 q115 q186 q187 q271

q102/control_k0/question.json

@@ -1,25 +1 @@
-{
-  "cell_id": "q102_control_k0",
-  "doc_id": "PS_2018.01.09_STEM_FINAL.pdf",
-  "question": "Among people in STEM jobs,  what is the type of STEP major with the largest gender gap consistent employment in the field that receive the highest pay in the survey of Pew Research Center, January 2018? ",
-  "answer": "computer science",
-  "answer_format": "Str",
-  "condition": "control",
-  "k": 0,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf"
-  ],
-  "pdf_names": [
-    "PS_2018.01.09_STEM_FINAL.pdf"
-  ],
-  "n_pages": 105,
-  "est_tokens": 94500,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf"
-  ],
-  "original_filenames": [
-    "PS_2018.01.09_STEM_FINAL.pdf"
-  ]
-}
+{"cell_id": "q102_control_k0", "doc_id": "PS_2018.01.09_STEM_FINAL.pdf", "question": "In the Pew Research Center survey (January 2018), by how many percentage points was the share of women in STEM who said they had experienced workplace discrimination higher than the share of women in non-STEM jobs who said the same?", "answer": "9", "answer_format": "Int", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf"], "pdf_names": ["PS_2018.01.09_STEM_FINAL.pdf"], "n_pages": 105, "est_tokens": 94500, "expected_route": "openrouter", "bundle_filenames": ["source.pdf"], "original_filenames": ["PS_2018.01.09_STEM_FINAL.pdf"]}

q102/hard_negative_k2/question.json

@@ -1 +1 @@
-{"cell_id": "q102_hard_negative_k2", "doc_id": "PS_2018.01.09_STEM_FINAL.pdf", "question": "Among people in STEM jobs,  what is the type of STEP major with the largest gender gap consistent employment in the field that receive the highest pay in the survey of Pew Research Center, January 2018? ", "answer": "computer science", "answer_format": "Str", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_d4c9f0dc3f6589cc.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/sheet_q102_pew_social.pdf"], "pdf_names": ["PS_2018.01.09_STEM_FINAL.pdf", "web_d4c9f0dc3f6589cc.pdf", "sheet_q102_pew_social.pdf"], "n_pages": 158, "est_tokens": 142200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["PS_2018.01.09_STEM_FINAL.pdf", "web_d4c9f0dc3f6589cc.pdf", "sheet_q102_pew_social.pdf"]}
+{"cell_id": "q102_hard_negative_k2", "doc_id": "PS_2018.01.09_STEM_FINAL.pdf", "question": "In the Pew Research Center survey (January 2018), by how many percentage points was the share of women in STEM who said they had experienced workplace discrimination higher than the share of women in non-STEM jobs who said the same?", "answer": "9", "answer_format": "Int", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_d4c9f0dc3f6589cc.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/sheet_q102_pew_social.pdf"], "pdf_names": ["PS_2018.01.09_STEM_FINAL.pdf", "web_d4c9f0dc3f6589cc.pdf", "sheet_q102_pew_social.pdf"], "n_pages": 158, "est_tokens": 142200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["PS_2018.01.09_STEM_FINAL.pdf", "web_d4c9f0dc3f6589cc.pdf", "sheet_q102_pew_social.pdf"]}

q102/hard_negative_k4/question.json

@@ -1 +1 @@
-{"cell_id": "q102_hard_negative_k4", "doc_id": "PS_2018.01.09_STEM_FINAL.pdf", "question": "Among people in STEM jobs,  what is the type of STEP major with the largest gender gap consistent employment in the field that receive the highest pay in the survey of Pew Research Center, January 2018? ", "answer": "computer science", "answer_format": "Str", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_d4c9f0dc3f6589cc.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/sheet_q102_pew_social.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_d3545b20cb487e4b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_9a644e720d72b089.pdf"], "pdf_names": ["PS_2018.01.09_STEM_FINAL.pdf", "web_d4c9f0dc3f6589cc.pdf", "sheet_q102_pew_social.pdf", "web_d3545b20cb487e4b.pdf", "web_9a644e720d72b089.pdf"], "n_pages": 229, "est_tokens": 206100, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["PS_2018.01.09_STEM_FINAL.pdf", "web_d4c9f0dc3f6589cc.pdf", "sheet_q102_pew_social.pdf", "web_d3545b20cb487e4b.pdf", "web_9a644e720d72b089.pdf"]}
+{"cell_id": "q102_hard_negative_k4", "doc_id": "PS_2018.01.09_STEM_FINAL.pdf", "question": "In the Pew Research Center survey (January 2018), by how many percentage points was the share of women in STEM who said they had experienced workplace discrimination higher than the share of women in non-STEM jobs who said the same?", "answer": "9", "answer_format": "Int", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_d4c9f0dc3f6589cc.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/sheet_q102_pew_social.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_d3545b20cb487e4b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_9a644e720d72b089.pdf"], "pdf_names": ["PS_2018.01.09_STEM_FINAL.pdf", "web_d4c9f0dc3f6589cc.pdf", "sheet_q102_pew_social.pdf", "web_d3545b20cb487e4b.pdf", "web_9a644e720d72b089.pdf"], "n_pages": 229, "est_tokens": 206100, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["PS_2018.01.09_STEM_FINAL.pdf", "web_d4c9f0dc3f6589cc.pdf", "sheet_q102_pew_social.pdf", "web_d3545b20cb487e4b.pdf", "web_9a644e720d72b089.pdf"]}

q102/random_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q102_random_k2",
-  "doc_id": "PS_2018.01.09_STEM_FINAL.pdf",
-  "question": "Among people in STEM jobs,  what is the type of STEP major with the largest gender gap consistent employment in the field that receive the highest pay in the survey of Pew Research Center, January 2018? ",
-  "answer": "computer science",
-  "answer_format": "Str",
-  "condition": "random",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_6c15b8b394b28edf.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_b873ee670af56beb.pdf"
-  ],
-  "pdf_names": [
-    "PS_2018.01.09_STEM_FINAL.pdf",
-    "web_6c15b8b394b28edf.pdf",
-    "web_b873ee670af56beb.pdf"
-  ],
-  "n_pages": 170,
-  "est_tokens": 153000,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf"
-  ],
-  "original_filenames": [
-    "PS_2018.01.09_STEM_FINAL.pdf",
-    "web_6c15b8b394b28edf.pdf",
-    "web_b873ee670af56beb.pdf"
-  ]
-}
+{"cell_id": "q102_random_k2", "doc_id": "PS_2018.01.09_STEM_FINAL.pdf", "question": "In the Pew Research Center survey (January 2018), by how many percentage points was the share of women in STEM who said they had experienced workplace discrimination higher than the share of women in non-STEM jobs who said the same?", "answer": "9", "answer_format": "Int", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_6c15b8b394b28edf.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_b873ee670af56beb.pdf"], "pdf_names": ["PS_2018.01.09_STEM_FINAL.pdf", "web_6c15b8b394b28edf.pdf", "web_b873ee670af56beb.pdf"], "n_pages": 170, "est_tokens": 153000, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["PS_2018.01.09_STEM_FINAL.pdf", "web_6c15b8b394b28edf.pdf", "web_b873ee670af56beb.pdf"]}

q102/random_k4/question.json

@@ -1,41 +1 @@
-{
-  "cell_id": "q102_random_k4",
-  "doc_id": "PS_2018.01.09_STEM_FINAL.pdf",
-  "question": "Among people in STEM jobs,  what is the type of STEP major with the largest gender gap consistent employment in the field that receive the highest pay in the survey of Pew Research Center, January 2018? ",
-  "answer": "computer science",
-  "answer_format": "Str",
-  "condition": "random",
-  "k": 4,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_6c15b8b394b28edf.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_b873ee670af56beb.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_25bb7468c3ed16b0.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_82328a4273fef79b.pdf"
-  ],
-  "pdf_names": [
-    "PS_2018.01.09_STEM_FINAL.pdf",
-    "web_6c15b8b394b28edf.pdf",
-    "web_b873ee670af56beb.pdf",
-    "web_25bb7468c3ed16b0.pdf",
-    "web_82328a4273fef79b.pdf"
-  ],
-  "n_pages": 227,
-  "est_tokens": 204300,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf",
-    "random_3.pdf",
-    "random_4.pdf"
-  ],
-  "original_filenames": [
-    "PS_2018.01.09_STEM_FINAL.pdf",
-    "web_6c15b8b394b28edf.pdf",
-    "web_b873ee670af56beb.pdf",
-    "web_25bb7468c3ed16b0.pdf",
-    "web_82328a4273fef79b.pdf"
-  ]
-}
+{"cell_id": "q102_random_k4", "doc_id": "PS_2018.01.09_STEM_FINAL.pdf", "question": "In the Pew Research Center survey (January 2018), by how many percentage points was the share of women in STEM who said they had experienced workplace discrimination higher than the share of women in non-STEM jobs who said the same?", "answer": "9", "answer_format": "Int", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/PS_2018.01.09_STEM_FINAL.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_6c15b8b394b28edf.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_b873ee670af56beb.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_25bb7468c3ed16b0.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_82328a4273fef79b.pdf"], "pdf_names": ["PS_2018.01.09_STEM_FINAL.pdf", "web_6c15b8b394b28edf.pdf", "web_b873ee670af56beb.pdf", "web_25bb7468c3ed16b0.pdf", "web_82328a4273fef79b.pdf"], "n_pages": 227, "est_tokens": 204300, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["PS_2018.01.09_STEM_FINAL.pdf", "web_6c15b8b394b28edf.pdf", "web_b873ee670af56beb.pdf", "web_25bb7468c3ed16b0.pdf", "web_82328a4273fef79b.pdf"]}

q104/control_k0/question.json

@@ -1 +1 @@
-{"cell_id": "q104_control_k0", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "How many committee members from Texas are presented in the conference of improving the quality and efficiency of commodity distribution to federal child nutrition programs?", "answer": "3", "answer_format": "Int", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf"], "n_pages": 20, "est_tokens": 18000, "expected_route": "azure", "bundle_filenames": ["source.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf"]}
+{"cell_id": "q104_control_k0", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "In the House committee whose roster is led by Chairman John A. Boehner of Ohio at the October 7, 2003 hearing, how many of its listed members represent Texas?", "answer": "3", "answer_format": "Int", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf"], "n_pages": 20, "est_tokens": 18000, "expected_route": "azure", "bundle_filenames": ["source.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf"]}

q104/hard_negative_k2/question.json

@@ -1 +1 @@
-{"cell_id": "q104_hard_negative_k2", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "How many committee members from Texas are presented in the conference of improving the quality and efficiency of commodity distribution to federal child nutrition programs?", "answer": "3", "answer_format": "Int", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_db15f36044519c81.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_6c15b8b394b28edf.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf", "web_db15f36044519c81.pdf", "web_6c15b8b394b28edf.pdf"], "n_pages": 88, "est_tokens": 79200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf", "web_db15f36044519c81.pdf", "web_6c15b8b394b28edf.pdf"]}
+{"cell_id": "q104_hard_negative_k2", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "In the House committee whose roster is led by Chairman John A. Boehner of Ohio at the October 7, 2003 hearing, how many of its listed members represent Texas?", "answer": "3", "answer_format": "Int", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_db15f36044519c81.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_6c15b8b394b28edf.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf", "web_db15f36044519c81.pdf", "web_6c15b8b394b28edf.pdf"], "n_pages": 88, "est_tokens": 79200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf", "web_db15f36044519c81.pdf", "web_6c15b8b394b28edf.pdf"]}

q104/hard_negative_k4/question.json

@@ -1 +1 @@
-{"cell_id": "q104_hard_negative_k4", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "How many committee members from Texas are presented in the conference of improving the quality and efficiency of commodity distribution to federal child nutrition programs?", "answer": "3", "answer_format": "Int", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_db15f36044519c81.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_6c15b8b394b28edf.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_7b4a136efb01e0d1.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_caed7cb4152d45e3.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf", "web_db15f36044519c81.pdf", "web_6c15b8b394b28edf.pdf", "web_7b4a136efb01e0d1.pdf", "web_caed7cb4152d45e3.pdf"], "n_pages": 144, "est_tokens": 129600, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf", "web_db15f36044519c81.pdf", "web_6c15b8b394b28edf.pdf", "web_7b4a136efb01e0d1.pdf", "web_caed7cb4152d45e3.pdf"]}
+{"cell_id": "q104_hard_negative_k4", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "In the House committee whose roster is led by Chairman John A. Boehner of Ohio at the October 7, 2003 hearing, how many of its listed members represent Texas?", "answer": "3", "answer_format": "Int", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_db15f36044519c81.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_6c15b8b394b28edf.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_7b4a136efb01e0d1.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_caed7cb4152d45e3.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf", "web_db15f36044519c81.pdf", "web_6c15b8b394b28edf.pdf", "web_7b4a136efb01e0d1.pdf", "web_caed7cb4152d45e3.pdf"], "n_pages": 144, "est_tokens": 129600, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf", "web_db15f36044519c81.pdf", "web_6c15b8b394b28edf.pdf", "web_7b4a136efb01e0d1.pdf", "web_caed7cb4152d45e3.pdf"]}

q104/random_k2/question.json

@@ -1 +1 @@
-{"cell_id": "q104_random_k2", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "How many committee members from Texas are presented in the conference of improving the quality and efficiency of commodity distribution to federal child nutrition programs?", "answer": "3", "answer_format": "Int", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0867a04cdab7a273.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_a39c0d7f5ff0dedb.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf", "web_0867a04cdab7a273.pdf", "web_a39c0d7f5ff0dedb.pdf"], "n_pages": 71, "est_tokens": 63900, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf", "web_0867a04cdab7a273.pdf", "web_a39c0d7f5ff0dedb.pdf"]}
+{"cell_id": "q104_random_k2", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "In the House committee whose roster is led by Chairman John A. Boehner of Ohio at the October 7, 2003 hearing, how many of its listed members represent Texas?", "answer": "3", "answer_format": "Int", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0867a04cdab7a273.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_a39c0d7f5ff0dedb.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf", "web_0867a04cdab7a273.pdf", "web_a39c0d7f5ff0dedb.pdf"], "n_pages": 71, "est_tokens": 63900, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf", "web_0867a04cdab7a273.pdf", "web_a39c0d7f5ff0dedb.pdf"]}

q104/random_k4/question.json

@@ -1 +1 @@
-{"cell_id": "q104_random_k4", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "How many committee members from Texas are presented in the conference of improving the quality and efficiency of commodity distribution to federal child nutrition programs?", "answer": "3", "answer_format": "Int", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0867a04cdab7a273.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_a39c0d7f5ff0dedb.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_5906edbb3f342646.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_660e5c500c425b81.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf", "web_0867a04cdab7a273.pdf", "web_a39c0d7f5ff0dedb.pdf", "web_5906edbb3f342646.pdf", "web_660e5c500c425b81.pdf"], "n_pages": 102, "est_tokens": 91800, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf", "web_0867a04cdab7a273.pdf", "web_a39c0d7f5ff0dedb.pdf", "web_5906edbb3f342646.pdf", "web_660e5c500c425b81.pdf"]}
+{"cell_id": "q104_random_k4", "doc_id": "e639029d16094ea71d964e2fb953952b.pdf", "question": "In the House committee whose roster is led by Chairman John A. Boehner of Ohio at the October 7, 2003 hearing, how many of its listed members represent Texas?", "answer": "3", "answer_format": "Int", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/e639029d16094ea71d964e2fb953952b.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0867a04cdab7a273.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_a39c0d7f5ff0dedb.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_5906edbb3f342646.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_660e5c500c425b81.pdf"], "pdf_names": ["e639029d16094ea71d964e2fb953952b.pdf", "web_0867a04cdab7a273.pdf", "web_a39c0d7f5ff0dedb.pdf", "web_5906edbb3f342646.pdf", "web_660e5c500c425b81.pdf"], "n_pages": 102, "est_tokens": 91800, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["e639029d16094ea71d964e2fb953952b.pdf", "web_0867a04cdab7a273.pdf", "web_a39c0d7f5ff0dedb.pdf", "web_5906edbb3f342646.pdf", "web_660e5c500c425b81.pdf"]}

q115/control_k0/question.json

@@ -1,25 +1 @@
-{
-  "cell_id": "q115_control_k0",
-  "doc_id": "2023.findings-emnlp.248.pdf",
-  "question": "Which dataset used in this paper was proposed in 2022 and all of its logical reasoning problems are multiple-choice questions with 5 options?",
-  "answer": "ProofWriter",
-  "answer_format": "Str",
-  "condition": "control",
-  "k": 0,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf"
-  ],
-  "pdf_names": [
-    "2023.findings-emnlp.248.pdf"
-  ],
-  "n_pages": 19,
-  "est_tokens": 17100,
-  "expected_route": "azure",
-  "bundle_filenames": [
-    "source.pdf"
-  ],
-  "original_filenames": [
-    "2023.findings-emnlp.248.pdf"
-  ]
-}
+{"cell_id": "q115_control_k0", "doc_id": "2023.findings-emnlp.248.pdf", "question": "In the paper main results table comparing Standard, Chain-of-Thought, and the proposed method across the five logical reasoning datasets, on which dataset does the proposed method (without self-refinement) achieve its highest accuracy when GPT-4 is the underlying LLM?", "answer": "LogicalDeduction", "answer_format": "Str", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf"], "pdf_names": ["2023.findings-emnlp.248.pdf"], "n_pages": 19, "est_tokens": 17100, "expected_route": "azure", "bundle_filenames": ["source.pdf"], "original_filenames": ["2023.findings-emnlp.248.pdf"]}

q115/hard_negative_k2/question.json

@@ -1 +1 @@
-{"cell_id": "q115_hard_negative_k2", "doc_id": "2023.findings-emnlp.248.pdf", "question": "Which dataset used in this paper was proposed in 2022 and all of its logical reasoning problems are multiple-choice questions with 5 options?", "answer": "ProofWriter", "answer_format": "Str", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_39b6cadc4e9ca03e.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_949ca008fd6aab55.pdf"], "pdf_names": ["2023.findings-emnlp.248.pdf", "web_39b6cadc4e9ca03e.pdf", "web_949ca008fd6aab55.pdf"], "n_pages": 58, "est_tokens": 52200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["2023.findings-emnlp.248.pdf", "web_39b6cadc4e9ca03e.pdf", "web_949ca008fd6aab55.pdf"]}
+{"cell_id": "q115_hard_negative_k2", "doc_id": "2023.findings-emnlp.248.pdf", "question": "In the paper main results table comparing Standard, Chain-of-Thought, and the proposed method across the five logical reasoning datasets, on which dataset does the proposed method (without self-refinement) achieve its highest accuracy when GPT-4 is the underlying LLM?", "answer": "LogicalDeduction", "answer_format": "Str", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_39b6cadc4e9ca03e.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_949ca008fd6aab55.pdf"], "pdf_names": ["2023.findings-emnlp.248.pdf", "web_39b6cadc4e9ca03e.pdf", "web_949ca008fd6aab55.pdf"], "n_pages": 58, "est_tokens": 52200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["2023.findings-emnlp.248.pdf", "web_39b6cadc4e9ca03e.pdf", "web_949ca008fd6aab55.pdf"]}

q115/hard_negative_k4/question.json

@@ -1 +1 @@
-{"cell_id": "q115_hard_negative_k4", "doc_id": "2023.findings-emnlp.248.pdf", "question": "Which dataset used in this paper was proposed in 2022 and all of its logical reasoning problems are multiple-choice questions with 5 options?", "answer": "ProofWriter", "answer_format": "Str", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_39b6cadc4e9ca03e.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_949ca008fd6aab55.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_56044d2e22a63da0.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_992a098addf13dd9.pdf"], "pdf_names": ["2023.findings-emnlp.248.pdf", "web_39b6cadc4e9ca03e.pdf", "web_949ca008fd6aab55.pdf", "web_56044d2e22a63da0.pdf", "web_992a098addf13dd9.pdf"], "n_pages": 107, "est_tokens": 96300, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["2023.findings-emnlp.248.pdf", "web_39b6cadc4e9ca03e.pdf", "web_949ca008fd6aab55.pdf", "web_56044d2e22a63da0.pdf", "web_992a098addf13dd9.pdf"]}
+{"cell_id": "q115_hard_negative_k4", "doc_id": "2023.findings-emnlp.248.pdf", "question": "In the paper main results table comparing Standard, Chain-of-Thought, and the proposed method across the five logical reasoning datasets, on which dataset does the proposed method (without self-refinement) achieve its highest accuracy when GPT-4 is the underlying LLM?", "answer": "LogicalDeduction", "answer_format": "Str", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_39b6cadc4e9ca03e.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_949ca008fd6aab55.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_56044d2e22a63da0.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_992a098addf13dd9.pdf"], "pdf_names": ["2023.findings-emnlp.248.pdf", "web_39b6cadc4e9ca03e.pdf", "web_949ca008fd6aab55.pdf", "web_56044d2e22a63da0.pdf", "web_992a098addf13dd9.pdf"], "n_pages": 107, "est_tokens": 96300, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["2023.findings-emnlp.248.pdf", "web_39b6cadc4e9ca03e.pdf", "web_949ca008fd6aab55.pdf", "web_56044d2e22a63da0.pdf", "web_992a098addf13dd9.pdf"]}

q115/random_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q115_random_k2",
-  "doc_id": "2023.findings-emnlp.248.pdf",
-  "question": "Which dataset used in this paper was proposed in 2022 and all of its logical reasoning problems are multiple-choice questions with 5 options?",
-  "answer": "ProofWriter",
-  "answer_format": "Str",
-  "condition": "random",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_7c476895c4da1bbe.pdf"
-  ],
-  "pdf_names": [
-    "2023.findings-emnlp.248.pdf",
-    "fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf",
-    "web_7c476895c4da1bbe.pdf"
-  ],
-  "n_pages": 46,
-  "est_tokens": 41400,
-  "expected_route": "azure",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf"
-  ],
-  "original_filenames": [
-    "2023.findings-emnlp.248.pdf",
-    "fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf",
-    "web_7c476895c4da1bbe.pdf"
-  ]
-}
+{"cell_id": "q115_random_k2", "doc_id": "2023.findings-emnlp.248.pdf", "question": "In the paper main results table comparing Standard, Chain-of-Thought, and the proposed method across the five logical reasoning datasets, on which dataset does the proposed method (without self-refinement) achieve its highest accuracy when GPT-4 is the underlying LLM?", "answer": "LogicalDeduction", "answer_format": "Str", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_7c476895c4da1bbe.pdf"], "pdf_names": ["2023.findings-emnlp.248.pdf", "fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf", "web_7c476895c4da1bbe.pdf"], "n_pages": 46, "est_tokens": 41400, "expected_route": "azure", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["2023.findings-emnlp.248.pdf", "fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf", "web_7c476895c4da1bbe.pdf"]}

q115/random_k4/question.json

@@ -1,41 +1 @@
-{
-  "cell_id": "q115_random_k4",
-  "doc_id": "2023.findings-emnlp.248.pdf",
-  "question": "Which dataset used in this paper was proposed in 2022 and all of its logical reasoning problems are multiple-choice questions with 5 options?",
-  "answer": "ProofWriter",
-  "answer_format": "Str",
-  "condition": "random",
-  "k": 4,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_7c476895c4da1bbe.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_86667079c4fbb42c.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0047db342bad4270.pdf"
-  ],
-  "pdf_names": [
-    "2023.findings-emnlp.248.pdf",
-    "fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf",
-    "web_7c476895c4da1bbe.pdf",
-    "web_86667079c4fbb42c.pdf",
-    "web_0047db342bad4270.pdf"
-  ],
-  "n_pages": 96,
-  "est_tokens": 86400,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf",
-    "random_3.pdf",
-    "random_4.pdf"
-  ],
-  "original_filenames": [
-    "2023.findings-emnlp.248.pdf",
-    "fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf",
-    "web_7c476895c4da1bbe.pdf",
-    "web_86667079c4fbb42c.pdf",
-    "web_0047db342bad4270.pdf"
-  ]
-}
+{"cell_id": "q115_random_k4", "doc_id": "2023.findings-emnlp.248.pdf", "question": "In the paper main results table comparing Standard, Chain-of-Thought, and the proposed method across the five logical reasoning datasets, on which dataset does the proposed method (without self-refinement) achieve its highest accuracy when GPT-4 is the underlying LLM?", "answer": "LogicalDeduction", "answer_format": "Str", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2023.findings-emnlp.248.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_7c476895c4da1bbe.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_86667079c4fbb42c.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0047db342bad4270.pdf"], "pdf_names": ["2023.findings-emnlp.248.pdf", "fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf", "web_7c476895c4da1bbe.pdf", "web_86667079c4fbb42c.pdf", "web_0047db342bad4270.pdf"], "n_pages": 96, "est_tokens": 86400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["2023.findings-emnlp.248.pdf", "fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf", "web_7c476895c4da1bbe.pdf", "web_86667079c4fbb42c.pdf", "web_0047db342bad4270.pdf"]}

q186/control_k0/question.json

@@ -1 +1 @@
-{"cell_id": "q186_control_k0", "doc_id": "2312.04350v3.pdf", "question": "What rung is the second ladder of causation refer to?", "answer": "intervention", "answer_format": "Str", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf"], "pdf_names": ["2312.04350v3.pdf"], "n_pages": 28, "est_tokens": 25200, "expected_route": "azure", "bundle_filenames": ["source.pdf"], "original_filenames": ["2312.04350v3.pdf"]}
+{"cell_id": "q186_control_k0", "doc_id": "2312.04350v3.pdf", "question": "In the chain-of-thought prompting strategy CAUSALCOT, by how many points does it improve over the performance of vanilla GPT-4 on the benchmark?", "answer": "8.37", "answer_format": "Float", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf"], "pdf_names": ["2312.04350v3.pdf"], "n_pages": 28, "est_tokens": 25200, "expected_route": "azure", "bundle_filenames": ["source.pdf"], "original_filenames": ["2312.04350v3.pdf"]}

q186/hard_negative_k2/question.json

@@ -1 +1 @@
-{"cell_id": "q186_hard_negative_k2", "doc_id": "2312.04350v3.pdf", "question": "What rung is the second ladder of causation refer to?", "answer": "intervention", "answer_format": "Str", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/arxiv_2404.06349.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/f556c55df76a9558422f00c4171900dc999c892a.pdf"], "pdf_names": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf"], "n_pages": 88, "est_tokens": 79200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf"]}
+{"cell_id": "q186_hard_negative_k2", "doc_id": "2312.04350v3.pdf", "question": "In the chain-of-thought prompting strategy CAUSALCOT, by how many points does it improve over the performance of vanilla GPT-4 on the benchmark?", "answer": "8.37", "answer_format": "Float", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/arxiv_2404.06349.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/f556c55df76a9558422f00c4171900dc999c892a.pdf"], "pdf_names": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf"], "n_pages": 88, "est_tokens": 79200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf"]}

q186/hard_negative_k4/question.json

@@ -1 +1 @@
-{"cell_id": "q186_hard_negative_k4", "doc_id": "2312.04350v3.pdf", "question": "What rung is the second ladder of causation refer to?", "answer": "intervention", "answer_format": "Str", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/arxiv_2404.06349.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/f556c55df76a9558422f00c4171900dc999c892a.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/sheet_q186_sighan.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"], "pdf_names": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf", "sheet_q186_sighan.pdf", "88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"], "n_pages": 108, "est_tokens": 97200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf", "sheet_q186_sighan.pdf", "88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"]}
+{"cell_id": "q186_hard_negative_k4", "doc_id": "2312.04350v3.pdf", "question": "In the chain-of-thought prompting strategy CAUSALCOT, by how many points does it improve over the performance of vanilla GPT-4 on the benchmark?", "answer": "8.37", "answer_format": "Float", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/arxiv_2404.06349.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/f556c55df76a9558422f00c4171900dc999c892a.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/sheet_q186_sighan.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"], "pdf_names": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf", "sheet_q186_sighan.pdf", "88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"], "n_pages": 108, "est_tokens": 97200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf", "sheet_q186_sighan.pdf", "88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"]}

q186/random_k2/question.json

@@ -1 +1 @@
-{"cell_id": "q186_random_k2", "doc_id": "2312.04350v3.pdf", "question": "What rung is the second ladder of causation refer to?", "answer": "intervention", "answer_format": "Str", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_6f67e9068ffec005.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_43df7d1ce2cae9ac.pdf"], "pdf_names": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf"], "n_pages": 84, "est_tokens": 75600, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf"]}
+{"cell_id": "q186_random_k2", "doc_id": "2312.04350v3.pdf", "question": "In the chain-of-thought prompting strategy CAUSALCOT, by how many points does it improve over the performance of vanilla GPT-4 on the benchmark?", "answer": "8.37", "answer_format": "Float", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_6f67e9068ffec005.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_43df7d1ce2cae9ac.pdf"], "pdf_names": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf"], "n_pages": 84, "est_tokens": 75600, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf"]}

q186/random_k4/question.json

@@ -1 +1 @@
-{"cell_id": "q186_random_k4", "doc_id": "2312.04350v3.pdf", "question": "What rung is the second ladder of causation refer to?", "answer": "intervention", "answer_format": "Str", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_6f67e9068ffec005.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_43df7d1ce2cae9ac.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0047db342bad4270.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_185a0d92e212a6ed.pdf"], "pdf_names": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf", "web_0047db342bad4270.pdf", "web_185a0d92e212a6ed.pdf"], "n_pages": 156, "est_tokens": 140400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf", "web_0047db342bad4270.pdf", "web_185a0d92e212a6ed.pdf"]}
+{"cell_id": "q186_random_k4", "doc_id": "2312.04350v3.pdf", "question": "In the chain-of-thought prompting strategy CAUSALCOT, by how many points does it improve over the performance of vanilla GPT-4 on the benchmark?", "answer": "8.37", "answer_format": "Float", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_6f67e9068ffec005.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_43df7d1ce2cae9ac.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0047db342bad4270.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_185a0d92e212a6ed.pdf"], "pdf_names": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf", "web_0047db342bad4270.pdf", "web_185a0d92e212a6ed.pdf"], "n_pages": 156, "est_tokens": 140400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf", "web_0047db342bad4270.pdf", "web_185a0d92e212a6ed.pdf"]}

q187/control_k0/question.json

@@ -1,25 +1 @@
-{
-  "cell_id": "q187_control_k0",
-  "doc_id": "2312.04350v3.pdf",
-  "question": "Which model performs the best on Cladder?",
-  "answer": "GPT-4",
-  "answer_format": "Int",
-  "condition": "control",
-  "k": 0,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf"
-  ],
-  "pdf_names": [
-    "2312.04350v3.pdf"
-  ],
-  "n_pages": 28,
-  "est_tokens": 25200,
-  "expected_route": "azure",
-  "bundle_filenames": [
-    "source.pdf"
-  ],
-  "original_filenames": [
-    "2312.04350v3.pdf"
-  ]
-}
+{"cell_id": "q187_control_k0", "doc_id": "2312.04350v3.pdf", "question": "In the main results table, how many accuracy points higher is the proposed CausalCoT chain-of-thought method than vanilla GPT-4 on Rung 1 (associational) questions? Give the difference to two decimal places.", "answer": "20.34", "answer_format": "Float", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf"], "pdf_names": ["2312.04350v3.pdf"], "n_pages": 28, "est_tokens": 25200, "expected_route": "azure", "bundle_filenames": ["source.pdf"], "original_filenames": ["2312.04350v3.pdf"]}

q187/hard_negative_k2/question.json

@@ -1 +1 @@
-{"cell_id": "q187_hard_negative_k2", "doc_id": "2312.04350v3.pdf", "question": "Which model performs the best on Cladder?", "answer": "GPT-4", "answer_format": "Int", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/arxiv_2404.06349.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/f556c55df76a9558422f00c4171900dc999c892a.pdf"], "pdf_names": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf"], "n_pages": 88, "est_tokens": 79200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf"]}
+{"cell_id": "q187_hard_negative_k2", "doc_id": "2312.04350v3.pdf", "question": "In the main results table, how many accuracy points higher is the proposed CausalCoT chain-of-thought method than vanilla GPT-4 on Rung 1 (associational) questions? Give the difference to two decimal places.", "answer": "20.34", "answer_format": "Float", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/arxiv_2404.06349.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/f556c55df76a9558422f00c4171900dc999c892a.pdf"], "pdf_names": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf"], "n_pages": 88, "est_tokens": 79200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf"]}

q187/hard_negative_k4/question.json

@@ -1 +1 @@
-{"cell_id": "q187_hard_negative_k4", "doc_id": "2312.04350v3.pdf", "question": "Which model performs the best on Cladder?", "answer": "GPT-4", "answer_format": "Int", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/arxiv_2404.06349.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/f556c55df76a9558422f00c4171900dc999c892a.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/sheet_q186_sighan.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"], "pdf_names": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf", "sheet_q186_sighan.pdf", "88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"], "n_pages": 108, "est_tokens": 97200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf", "sheet_q186_sighan.pdf", "88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"]}
+{"cell_id": "q187_hard_negative_k4", "doc_id": "2312.04350v3.pdf", "question": "In the main results table, how many accuracy points higher is the proposed CausalCoT chain-of-thought method than vanilla GPT-4 on Rung 1 (associational) questions? Give the difference to two decimal places.", "answer": "20.34", "answer_format": "Float", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/arxiv_2404.06349.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/f556c55df76a9558422f00c4171900dc999c892a.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/sheet_q186_sighan.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"], "pdf_names": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf", "sheet_q186_sighan.pdf", "88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"], "n_pages": 108, "est_tokens": 97200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["2312.04350v3.pdf", "arxiv_2404.06349.pdf", "f556c55df76a9558422f00c4171900dc999c892a.pdf", "sheet_q186_sighan.pdf", "88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf"]}

q187/random_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q187_random_k2",
-  "doc_id": "2312.04350v3.pdf",
-  "question": "Which model performs the best on Cladder?",
-  "answer": "GPT-4",
-  "answer_format": "Int",
-  "condition": "random",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_6f67e9068ffec005.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_43df7d1ce2cae9ac.pdf"
-  ],
-  "pdf_names": [
-    "2312.04350v3.pdf",
-    "web_6f67e9068ffec005.pdf",
-    "web_43df7d1ce2cae9ac.pdf"
-  ],
-  "n_pages": 84,
-  "est_tokens": 75600,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf"
-  ],
-  "original_filenames": [
-    "2312.04350v3.pdf",
-    "web_6f67e9068ffec005.pdf",
-    "web_43df7d1ce2cae9ac.pdf"
-  ]
-}
+{"cell_id": "q187_random_k2", "doc_id": "2312.04350v3.pdf", "question": "In the main results table, how many accuracy points higher is the proposed CausalCoT chain-of-thought method than vanilla GPT-4 on Rung 1 (associational) questions? Give the difference to two decimal places.", "answer": "20.34", "answer_format": "Float", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_6f67e9068ffec005.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_43df7d1ce2cae9ac.pdf"], "pdf_names": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf"], "n_pages": 84, "est_tokens": 75600, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf"]}

q187/random_k4/question.json

@@ -1,41 +1 @@
-{
-  "cell_id": "q187_random_k4",
-  "doc_id": "2312.04350v3.pdf",
-  "question": "Which model performs the best on Cladder?",
-  "answer": "GPT-4",
-  "answer_format": "Int",
-  "condition": "random",
-  "k": 4,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_6f67e9068ffec005.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_43df7d1ce2cae9ac.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0047db342bad4270.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_185a0d92e212a6ed.pdf"
-  ],
-  "pdf_names": [
-    "2312.04350v3.pdf",
-    "web_6f67e9068ffec005.pdf",
-    "web_43df7d1ce2cae9ac.pdf",
-    "web_0047db342bad4270.pdf",
-    "web_185a0d92e212a6ed.pdf"
-  ],
-  "n_pages": 156,
-  "est_tokens": 140400,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf",
-    "random_3.pdf",
-    "random_4.pdf"
-  ],
-  "original_filenames": [
-    "2312.04350v3.pdf",
-    "web_6f67e9068ffec005.pdf",
-    "web_43df7d1ce2cae9ac.pdf",
-    "web_0047db342bad4270.pdf",
-    "web_185a0d92e212a6ed.pdf"
-  ]
-}
+{"cell_id": "q187_random_k4", "doc_id": "2312.04350v3.pdf", "question": "In the main results table, how many accuracy points higher is the proposed CausalCoT chain-of-thought method than vanilla GPT-4 on Rung 1 (associational) questions? Give the difference to two decimal places.", "answer": "20.34", "answer_format": "Float", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/2312.04350v3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_6f67e9068ffec005.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_43df7d1ce2cae9ac.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_0047db342bad4270.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_185a0d92e212a6ed.pdf"], "pdf_names": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf", "web_0047db342bad4270.pdf", "web_185a0d92e212a6ed.pdf"], "n_pages": 156, "est_tokens": 140400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["2312.04350v3.pdf", "web_6f67e9068ffec005.pdf", "web_43df7d1ce2cae9ac.pdf", "web_0047db342bad4270.pdf", "web_185a0d92e212a6ed.pdf"]}

q271/control_k0/question.json

@@ -1,25 +1 @@
-{
-  "cell_id": "q271_control_k0",
-  "doc_id": "BESTBUY_2023_10K.pdf",
-  "question": "what is interest coverage ratio for AMCOR'FY 2020? round your answer to three decimal?",
-  "answer": "51.286",
-  "answer_format": "Float",
-  "condition": "control",
-  "k": 0,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf"
-  ],
-  "pdf_names": [
-    "BESTBUY_2023_10K.pdf"
-  ],
-  "n_pages": 75,
-  "est_tokens": 67500,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf"
-  ],
-  "original_filenames": [
-    "BESTBUY_2023_10K.pdf"
-  ]
-}
+{"cell_id": "q271_control_k0", "doc_id": "BESTBUY_2023_10K.pdf", "question": "What is the interest coverage ratio (operating income divided by interest expense) for Best Buy for the fiscal year ended January 28, 2023? Round your answer to three decimals.", "answer": "51.286", "answer_format": "Float", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf"], "pdf_names": ["BESTBUY_2023_10K.pdf"], "n_pages": 75, "est_tokens": 67500, "expected_route": "openrouter", "bundle_filenames": ["source.pdf"], "original_filenames": ["BESTBUY_2023_10K.pdf"]}

q271/hard_negative_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q271_hard_negative_k2",
-  "doc_id": "BESTBUY_2023_10K.pdf",
-  "question": "what is interest coverage ratio for AMCOR'FY 2020? round your answer to three decimal?",
-  "answer": "51.286",
-  "answer_format": "Float",
-  "condition": "hard_negative",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_88be333a901981bb.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_839cc267cf5fcc63.pdf"
-  ],
-  "pdf_names": [
-    "BESTBUY_2023_10K.pdf",
-    "web_88be333a901981bb.pdf",
-    "web_839cc267cf5fcc63.pdf"
-  ],
-  "n_pages": 106,
-  "est_tokens": 95400,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "hn_1.pdf",
-    "hn_2.pdf"
-  ],
-  "original_filenames": [
-    "BESTBUY_2023_10K.pdf",
-    "web_88be333a901981bb.pdf",
-    "web_839cc267cf5fcc63.pdf"
-  ]
-}
+{"cell_id": "q271_hard_negative_k2", "doc_id": "BESTBUY_2023_10K.pdf", "question": "What is the interest coverage ratio (operating income divided by interest expense) for Best Buy for the fiscal year ended January 28, 2023? Round your answer to three decimals.", "answer": "51.286", "answer_format": "Float", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_88be333a901981bb.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_839cc267cf5fcc63.pdf"], "pdf_names": ["BESTBUY_2023_10K.pdf", "web_88be333a901981bb.pdf", "web_839cc267cf5fcc63.pdf"], "n_pages": 106, "est_tokens": 95400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["BESTBUY_2023_10K.pdf", "web_88be333a901981bb.pdf", "web_839cc267cf5fcc63.pdf"]}

q271/hard_negative_k4/question.json

@@ -1 +1 @@
-{"cell_id": "q271_hard_negative_k4", "doc_id": "BESTBUY_2023_10K.pdf", "question": "what is interest coverage ratio for AMCOR'FY 2020? round your answer to three decimal?", "answer": "51.286", "answer_format": "Float", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_88be333a901981bb.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_839cc267cf5fcc63.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_cd0ef8929be1ba45.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_4eb4afea574790a6.pdf"], "pdf_names": ["BESTBUY_2023_10K.pdf", "web_88be333a901981bb.pdf", "web_839cc267cf5fcc63.pdf", "web_cd0ef8929be1ba45.pdf", "web_4eb4afea574790a6.pdf"], "n_pages": 286, "est_tokens": 257400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["BESTBUY_2023_10K.pdf", "web_88be333a901981bb.pdf", "web_839cc267cf5fcc63.pdf", "web_cd0ef8929be1ba45.pdf", "web_4eb4afea574790a6.pdf"]}
+{"cell_id": "q271_hard_negative_k4", "doc_id": "BESTBUY_2023_10K.pdf", "question": "What is the interest coverage ratio (operating income divided by interest expense) for Best Buy for the fiscal year ended January 28, 2023? Round your answer to three decimals.", "answer": "51.286", "answer_format": "Float", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_88be333a901981bb.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_839cc267cf5fcc63.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_cd0ef8929be1ba45.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_4eb4afea574790a6.pdf"], "pdf_names": ["BESTBUY_2023_10K.pdf", "web_88be333a901981bb.pdf", "web_839cc267cf5fcc63.pdf", "web_cd0ef8929be1ba45.pdf", "web_4eb4afea574790a6.pdf"], "n_pages": 286, "est_tokens": 257400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["BESTBUY_2023_10K.pdf", "web_88be333a901981bb.pdf", "web_839cc267cf5fcc63.pdf", "web_cd0ef8929be1ba45.pdf", "web_4eb4afea574790a6.pdf"]}

q271/random_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q271_random_k2",
-  "doc_id": "BESTBUY_2023_10K.pdf",
-  "question": "what is interest coverage ratio for AMCOR'FY 2020? round your answer to three decimal?",
-  "answer": "51.286",
-  "answer_format": "Float",
-  "condition": "random",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_459cb41b80747aec.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf"
-  ],
-  "pdf_names": [
-    "BESTBUY_2023_10K.pdf",
-    "web_459cb41b80747aec.pdf",
-    "eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf"
-  ],
-  "n_pages": 106,
-  "est_tokens": 95400,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf"
-  ],
-  "original_filenames": [
-    "BESTBUY_2023_10K.pdf",
-    "web_459cb41b80747aec.pdf",
-    "eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf"
-  ]
-}
+{"cell_id": "q271_random_k2", "doc_id": "BESTBUY_2023_10K.pdf", "question": "What is the interest coverage ratio (operating income divided by interest expense) for Best Buy for the fiscal year ended January 28, 2023? Round your answer to three decimals.", "answer": "51.286", "answer_format": "Float", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_459cb41b80747aec.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf"], "pdf_names": ["BESTBUY_2023_10K.pdf", "web_459cb41b80747aec.pdf", "eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf"], "n_pages": 106, "est_tokens": 95400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["BESTBUY_2023_10K.pdf", "web_459cb41b80747aec.pdf", "eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf"]}

q271/random_k4/question.json

@@ -1,41 +1 @@
-{
-  "cell_id": "q271_random_k4",
-  "doc_id": "BESTBUY_2023_10K.pdf",
-  "question": "what is interest coverage ratio for AMCOR'FY 2020? round your answer to three decimal?",
-  "answer": "51.286",
-  "answer_format": "Float",
-  "condition": "random",
-  "k": 4,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_459cb41b80747aec.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_715e6cf2d4ba3d94.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_c916dc9b2089aeec.pdf"
-  ],
-  "pdf_names": [
-    "BESTBUY_2023_10K.pdf",
-    "web_459cb41b80747aec.pdf",
-    "eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf",
-    "web_715e6cf2d4ba3d94.pdf",
-    "web_c916dc9b2089aeec.pdf"
-  ],
-  "n_pages": 146,
-  "est_tokens": 131400,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf",
-    "random_3.pdf",
-    "random_4.pdf"
-  ],
-  "original_filenames": [
-    "BESTBUY_2023_10K.pdf",
-    "web_459cb41b80747aec.pdf",
-    "eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf",
-    "web_715e6cf2d4ba3d94.pdf",
-    "web_c916dc9b2089aeec.pdf"
-  ]
-}
+{"cell_id": "q271_random_k4", "doc_id": "BESTBUY_2023_10K.pdf", "question": "What is the interest coverage ratio (operating income divided by interest expense) for Best Buy for the fiscal year ended January 28, 2023? Round your answer to three decimals.", "answer": "51.286", "answer_format": "Float", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/BESTBUY_2023_10K.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_459cb41b80747aec.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_715e6cf2d4ba3d94.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_c916dc9b2089aeec.pdf"], "pdf_names": ["BESTBUY_2023_10K.pdf", "web_459cb41b80747aec.pdf", "eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf", "web_715e6cf2d4ba3d94.pdf", "web_c916dc9b2089aeec.pdf"], "n_pages": 146, "est_tokens": 131400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["BESTBUY_2023_10K.pdf", "web_459cb41b80747aec.pdf", "eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf", "web_715e6cf2d4ba3d94.pdf", "web_c916dc9b2089aeec.pdf"]}

q64/control_k0/question.json

@@ -1,25 +1 @@
-{
-  "cell_id": "q64_control_k0",
-  "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-  "question": "A transmitter operating at a frequency of 500 MHz has a rated maximum output power of 10 W. Using the recommended separation distances between portable and mobile RF communications equipment and the M270TF-XXX / M320TF-XXX, calculate the minimum separation distance required for this transmitter. Write down the answer in float format with 2 decimals.",
-  "answer": "2.92",
-  "answer_format": "Float",
-  "condition": "control",
-  "k": 0,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf"
-  ],
-  "pdf_names": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf"
-  ],
-  "n_pages": 40,
-  "est_tokens": 36000,
-  "expected_route": "azure",
-  "bundle_filenames": [
-    "source.pdf"
-  ],
-  "original_filenames": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf"
-  ]
-}
+{"cell_id": "q64_control_k0", "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "question": "According to the Technical Specification appendix of the 4K UHD Display User Manual, what is the horizontal pixel pitch (in micrometers) of the M320TF-XX monitor model?", "answer": "184.5", "answer_format": "Float", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf"], "pdf_names": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf"], "n_pages": 40, "est_tokens": 36000, "expected_route": "azure", "bundle_filenames": ["source.pdf"], "original_filenames": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf"]}

q64/hard_negative_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q64_hard_negative_k2",
-  "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-  "question": "A transmitter operating at a frequency of 500 MHz has a rated maximum output power of 10 W. Using the recommended separation distances between portable and mobile RF communications equipment and the M270TF-XXX / M320TF-XXX, calculate the minimum separation distance required for this transmitter. Write down the answer in float format with 2 decimals.",
-  "answer": "2.92",
-  "answer_format": "Float",
-  "condition": "hard_negative",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_8bce5014fb268e2e.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_ca82e89a96a0fe65.pdf"
-  ],
-  "pdf_names": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "web_8bce5014fb268e2e.pdf",
-    "web_ca82e89a96a0fe65.pdf"
-  ],
-  "n_pages": 108,
-  "est_tokens": 97200,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "hn_1.pdf",
-    "hn_2.pdf"
-  ],
-  "original_filenames": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "web_8bce5014fb268e2e.pdf",
-    "web_ca82e89a96a0fe65.pdf"
-  ]
-}
+{"cell_id": "q64_hard_negative_k2", "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "question": "According to the Technical Specification appendix of the 4K UHD Display User Manual, what is the horizontal pixel pitch (in micrometers) of the M320TF-XX monitor model?", "answer": "184.5", "answer_format": "Float", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_8bce5014fb268e2e.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_ca82e89a96a0fe65.pdf"], "pdf_names": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "web_8bce5014fb268e2e.pdf", "web_ca82e89a96a0fe65.pdf"], "n_pages": 108, "est_tokens": 97200, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "web_8bce5014fb268e2e.pdf", "web_ca82e89a96a0fe65.pdf"]}

q64/hard_negative_k4/question.json

@@ -1,41 +1 @@
-{
-  "cell_id": "q64_hard_negative_k4",
-  "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-  "question": "A transmitter operating at a frequency of 500 MHz has a rated maximum output power of 10 W. Using the recommended separation distances between portable and mobile RF communications equipment and the M270TF-XXX / M320TF-XXX, calculate the minimum separation distance required for this transmitter. Write down the answer in float format with 2 decimals.",
-  "answer": "2.92",
-  "answer_format": "Float",
-  "condition": "hard_negative",
-  "k": 4,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_8bce5014fb268e2e.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_ca82e89a96a0fe65.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_f539c13fe18de9dd.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_1fed1ad9de3ee391.pdf"
-  ],
-  "pdf_names": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "web_8bce5014fb268e2e.pdf",
-    "web_ca82e89a96a0fe65.pdf",
-    "web_f539c13fe18de9dd.pdf",
-    "web_1fed1ad9de3ee391.pdf"
-  ],
-  "n_pages": 189,
-  "est_tokens": 170100,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "hn_1.pdf",
-    "hn_2.pdf",
-    "hn_3.pdf",
-    "hn_4.pdf"
-  ],
-  "original_filenames": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "web_8bce5014fb268e2e.pdf",
-    "web_ca82e89a96a0fe65.pdf",
-    "web_f539c13fe18de9dd.pdf",
-    "web_1fed1ad9de3ee391.pdf"
-  ]
-}
+{"cell_id": "q64_hard_negative_k4", "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "question": "According to the Technical Specification appendix of the 4K UHD Display User Manual, what is the horizontal pixel pitch (in micrometers) of the M320TF-XX monitor model?", "answer": "184.5", "answer_format": "Float", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_8bce5014fb268e2e.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_ca82e89a96a0fe65.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_f539c13fe18de9dd.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_1fed1ad9de3ee391.pdf"], "pdf_names": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "web_8bce5014fb268e2e.pdf", "web_ca82e89a96a0fe65.pdf", "web_f539c13fe18de9dd.pdf", "web_1fed1ad9de3ee391.pdf"], "n_pages": 189, "est_tokens": 170100, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "web_8bce5014fb268e2e.pdf", "web_ca82e89a96a0fe65.pdf", "web_f539c13fe18de9dd.pdf", "web_1fed1ad9de3ee391.pdf"]}

q64/random_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q64_random_k2",
-  "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-  "question": "A transmitter operating at a frequency of 500 MHz has a rated maximum output power of 10 W. Using the recommended separation distances between portable and mobile RF communications equipment and the M270TF-XXX / M320TF-XXX, calculate the minimum separation distance required for this transmitter. Write down the answer in float format with 2 decimals.",
-  "answer": "2.92",
-  "answer_format": "Float",
-  "condition": "random",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_d0752a6a77ae08fd.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_2be6edaf3332034d.pdf"
-  ],
-  "pdf_names": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "web_d0752a6a77ae08fd.pdf",
-    "web_2be6edaf3332034d.pdf"
-  ],
-  "n_pages": 91,
-  "est_tokens": 81900,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf"
-  ],
-  "original_filenames": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "web_d0752a6a77ae08fd.pdf",
-    "web_2be6edaf3332034d.pdf"
-  ]
-}
+{"cell_id": "q64_random_k2", "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "question": "According to the Technical Specification appendix of the 4K UHD Display User Manual, what is the horizontal pixel pitch (in micrometers) of the M320TF-XX monitor model?", "answer": "184.5", "answer_format": "Float", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_d0752a6a77ae08fd.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_2be6edaf3332034d.pdf"], "pdf_names": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "web_d0752a6a77ae08fd.pdf", "web_2be6edaf3332034d.pdf"], "n_pages": 91, "est_tokens": 81900, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "web_d0752a6a77ae08fd.pdf", "web_2be6edaf3332034d.pdf"]}

q64/random_k4/question.json

@@ -1,41 +1 @@
-{
-  "cell_id": "q64_random_k4",
-  "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-  "question": "A transmitter operating at a frequency of 500 MHz has a rated maximum output power of 10 W. Using the recommended separation distances between portable and mobile RF communications equipment and the M270TF-XXX / M320TF-XXX, calculate the minimum separation distance required for this transmitter. Write down the answer in float format with 2 decimals.",
-  "answer": "2.92",
-  "answer_format": "Float",
-  "condition": "random",
-  "k": 4,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_d0752a6a77ae08fd.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_2be6edaf3332034d.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_2ccba6013fe960b7.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/c726c35b04831158d54cee8acd7202672bb42877.pdf"
-  ],
-  "pdf_names": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "web_d0752a6a77ae08fd.pdf",
-    "web_2be6edaf3332034d.pdf",
-    "web_2ccba6013fe960b7.pdf",
-    "c726c35b04831158d54cee8acd7202672bb42877.pdf"
-  ],
-  "n_pages": 115,
-  "est_tokens": 103500,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf",
-    "random_3.pdf",
-    "random_4.pdf"
-  ],
-  "original_filenames": [
-    "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf",
-    "web_d0752a6a77ae08fd.pdf",
-    "web_2be6edaf3332034d.pdf",
-    "web_2ccba6013fe960b7.pdf",
-    "c726c35b04831158d54cee8acd7202672bb42877.pdf"
-  ]
-}
+{"cell_id": "q64_random_k4", "doc_id": "91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "question": "According to the Technical Specification appendix of the 4K UHD Display User Manual, what is the horizontal pixel pitch (in micrometers) of the M320TF-XX monitor model?", "answer": "184.5", "answer_format": "Float", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_d0752a6a77ae08fd.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_2be6edaf3332034d.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_2ccba6013fe960b7.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/c726c35b04831158d54cee8acd7202672bb42877.pdf"], "pdf_names": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "web_d0752a6a77ae08fd.pdf", "web_2be6edaf3332034d.pdf", "web_2ccba6013fe960b7.pdf", "c726c35b04831158d54cee8acd7202672bb42877.pdf"], "n_pages": 115, "est_tokens": 103500, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf", "web_d0752a6a77ae08fd.pdf", "web_2be6edaf3332034d.pdf", "web_2ccba6013fe960b7.pdf", "c726c35b04831158d54cee8acd7202672bb42877.pdf"]}

q85/control_k0/question.json

@@ -1,25 +1 @@
-{
-  "cell_id": "q85_control_k0",
-  "doc_id": "SnapNTell.pdf",
-  "question": "In the 3rd Wiki filtering, how many more entities were filtered out compared to the 1st Wiki filtering?",
-  "answer": "2885",
-  "answer_format": "Int",
-  "condition": "control",
-  "k": 0,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf"
-  ],
-  "pdf_names": [
-    "SnapNTell.pdf"
-  ],
-  "n_pages": 20,
-  "est_tokens": 18000,
-  "expected_route": "azure",
-  "bundle_filenames": [
-    "source.pdf"
-  ],
-  "original_filenames": [
-    "SnapNTell.pdf"
-  ]
-}
+{"cell_id": "q85_control_k0", "doc_id": "SnapNTell.pdf", "question": "According to the entity dataset filtering statistics, how many entities were removed by the 1st Wiki filtering stage (which removes entities without a wiki page)?", "answer": "4457", "answer_format": "Int", "condition": "control", "k": 0, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf"], "pdf_names": ["SnapNTell.pdf"], "n_pages": 20, "est_tokens": 18000, "expected_route": "azure", "bundle_filenames": ["source.pdf"], "original_filenames": ["SnapNTell.pdf"]}

q85/hard_negative_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q85_hard_negative_k2",
-  "doc_id": "SnapNTell.pdf",
-  "question": "In the 3rd Wiki filtering, how many more entities were filtered out compared to the 1st Wiki filtering?",
-  "answer": "2885",
-  "answer_format": "Int",
-  "condition": "hard_negative",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_4dcbbf87758697e3.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_99fd395a0c29c14c.pdf"
-  ],
-  "pdf_names": [
-    "SnapNTell.pdf",
-    "web_4dcbbf87758697e3.pdf",
-    "web_99fd395a0c29c14c.pdf"
-  ],
-  "n_pages": 48,
-  "est_tokens": 43200,
-  "expected_route": "azure",
-  "bundle_filenames": [
-    "source.pdf",
-    "hn_1.pdf",
-    "hn_2.pdf"
-  ],
-  "original_filenames": [
-    "SnapNTell.pdf",
-    "web_4dcbbf87758697e3.pdf",
-    "web_99fd395a0c29c14c.pdf"
-  ]
-}
+{"cell_id": "q85_hard_negative_k2", "doc_id": "SnapNTell.pdf", "question": "According to the entity dataset filtering statistics, how many entities were removed by the 1st Wiki filtering stage (which removes entities without a wiki page)?", "answer": "4457", "answer_format": "Int", "condition": "hard_negative", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_4dcbbf87758697e3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_99fd395a0c29c14c.pdf"], "pdf_names": ["SnapNTell.pdf", "web_4dcbbf87758697e3.pdf", "web_99fd395a0c29c14c.pdf"], "n_pages": 48, "est_tokens": 43200, "expected_route": "azure", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf"], "original_filenames": ["SnapNTell.pdf", "web_4dcbbf87758697e3.pdf", "web_99fd395a0c29c14c.pdf"]}

q85/hard_negative_k4/question.json

@@ -1,41 +1 @@
-{
-  "cell_id": "q85_hard_negative_k4",
-  "doc_id": "SnapNTell.pdf",
-  "question": "In the 3rd Wiki filtering, how many more entities were filtered out compared to the 1st Wiki filtering?",
-  "answer": "2885",
-  "answer_format": "Int",
-  "condition": "hard_negative",
-  "k": 4,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_4dcbbf87758697e3.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_99fd395a0c29c14c.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_87a18654e37349f2.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_903701db648fd6ad.pdf"
-  ],
-  "pdf_names": [
-    "SnapNTell.pdf",
-    "web_4dcbbf87758697e3.pdf",
-    "web_99fd395a0c29c14c.pdf",
-    "web_87a18654e37349f2.pdf",
-    "web_903701db648fd6ad.pdf"
-  ],
-  "n_pages": 86,
-  "est_tokens": 77400,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "hn_1.pdf",
-    "hn_2.pdf",
-    "hn_3.pdf",
-    "hn_4.pdf"
-  ],
-  "original_filenames": [
-    "SnapNTell.pdf",
-    "web_4dcbbf87758697e3.pdf",
-    "web_99fd395a0c29c14c.pdf",
-    "web_87a18654e37349f2.pdf",
-    "web_903701db648fd6ad.pdf"
-  ]
-}
+{"cell_id": "q85_hard_negative_k4", "doc_id": "SnapNTell.pdf", "question": "According to the entity dataset filtering statistics, how many entities were removed by the 1st Wiki filtering stage (which removes entities without a wiki page)?", "answer": "4457", "answer_format": "Int", "condition": "hard_negative", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_4dcbbf87758697e3.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_99fd395a0c29c14c.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_87a18654e37349f2.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_903701db648fd6ad.pdf"], "pdf_names": ["SnapNTell.pdf", "web_4dcbbf87758697e3.pdf", "web_99fd395a0c29c14c.pdf", "web_87a18654e37349f2.pdf", "web_903701db648fd6ad.pdf"], "n_pages": 86, "est_tokens": 77400, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "hn_1.pdf", "hn_2.pdf", "hn_3.pdf", "hn_4.pdf"], "original_filenames": ["SnapNTell.pdf", "web_4dcbbf87758697e3.pdf", "web_99fd395a0c29c14c.pdf", "web_87a18654e37349f2.pdf", "web_903701db648fd6ad.pdf"]}

q85/random_k2/question.json

@@ -1,33 +1 @@
-{
-  "cell_id": "q85_random_k2",
-  "doc_id": "SnapNTell.pdf",
-  "question": "In the 3rd Wiki filtering, how many more entities were filtered out compared to the 1st Wiki filtering?",
-  "answer": "2885",
-  "answer_format": "Int",
-  "condition": "random",
-  "k": 2,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_88be333a901981bb.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_18e98cd838edeb2e.pdf"
-  ],
-  "pdf_names": [
-    "SnapNTell.pdf",
-    "web_88be333a901981bb.pdf",
-    "web_18e98cd838edeb2e.pdf"
-  ],
-  "n_pages": 52,
-  "est_tokens": 46800,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf"
-  ],
-  "original_filenames": [
-    "SnapNTell.pdf",
-    "web_88be333a901981bb.pdf",
-    "web_18e98cd838edeb2e.pdf"
-  ]
-}
+{"cell_id": "q85_random_k2", "doc_id": "SnapNTell.pdf", "question": "According to the entity dataset filtering statistics, how many entities were removed by the 1st Wiki filtering stage (which removes entities without a wiki page)?", "answer": "4457", "answer_format": "Int", "condition": "random", "k": 2, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_88be333a901981bb.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_18e98cd838edeb2e.pdf"], "pdf_names": ["SnapNTell.pdf", "web_88be333a901981bb.pdf", "web_18e98cd838edeb2e.pdf"], "n_pages": 52, "est_tokens": 46800, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf"], "original_filenames": ["SnapNTell.pdf", "web_88be333a901981bb.pdf", "web_18e98cd838edeb2e.pdf"]}

q85/random_k4/question.json

@@ -1,41 +1 @@
-{
-  "cell_id": "q85_random_k4",
-  "doc_id": "SnapNTell.pdf",
-  "question": "In the 3rd Wiki filtering, how many more entities were filtered out compared to the 1st Wiki filtering?",
-  "answer": "2885",
-  "answer_format": "Int",
-  "condition": "random",
-  "k": 4,
-  "source_position": "first",
-  "pdf_paths": [
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_88be333a901981bb.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_18e98cd838edeb2e.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_3dbc7a33416787aa.pdf",
-    "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_ff408b2c8a1714b8.pdf"
-  ],
-  "pdf_names": [
-    "SnapNTell.pdf",
-    "web_88be333a901981bb.pdf",
-    "web_18e98cd838edeb2e.pdf",
-    "web_3dbc7a33416787aa.pdf",
-    "web_ff408b2c8a1714b8.pdf"
-  ],
-  "n_pages": 107,
-  "est_tokens": 96300,
-  "expected_route": "openrouter",
-  "bundle_filenames": [
-    "source.pdf",
-    "random_1.pdf",
-    "random_2.pdf",
-    "random_3.pdf",
-    "random_4.pdf"
-  ],
-  "original_filenames": [
-    "SnapNTell.pdf",
-    "web_88be333a901981bb.pdf",
-    "web_18e98cd838edeb2e.pdf",
-    "web_3dbc7a33416787aa.pdf",
-    "web_ff408b2c8a1714b8.pdf"
-  ]
-}
+{"cell_id": "q85_random_k4", "doc_id": "SnapNTell.pdf", "question": "According to the entity dataset filtering statistics, how many entities were removed by the 1st Wiki filtering stage (which removes entities without a wiki page)?", "answer": "4457", "answer_format": "Int", "condition": "random", "k": 4, "source_position": "first", "pdf_paths": ["/Users/jasonluo08/Desktop/AI Brain/algoverse_research/pdfs/SnapNTell.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs_resized/web_88be333a901981bb.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_18e98cd838edeb2e.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_3dbc7a33416787aa.pdf", "/Users/jasonluo08/Desktop/AI Brain/algoverse_research/hard_negative_pdfs/web_ff408b2c8a1714b8.pdf"], "pdf_names": ["SnapNTell.pdf", "web_88be333a901981bb.pdf", "web_18e98cd838edeb2e.pdf", "web_3dbc7a33416787aa.pdf", "web_ff408b2c8a1714b8.pdf"], "n_pages": 107, "est_tokens": 96300, "expected_route": "openrouter", "bundle_filenames": ["source.pdf", "random_1.pdf", "random_2.pdf", "random_3.pdf", "random_4.pdf"], "original_filenames": ["SnapNTell.pdf", "web_88be333a901981bb.pdf", "web_18e98cd838edeb2e.pdf", "web_3dbc7a33416787aa.pdf", "web_ff408b2c8a1714b8.pdf"]}

questions.csv

@@ -63,7 +63,7 @@ q60,2310.07609v1.pdf,What is the average performance gap between the ProgramFC p
 q61,2303.05039v2.pdf,How many items in total from Amazon-related datasets were used in the paper that made two new personality datasets?,8980,Int,2303.05039v2.pdf,9,8125d1dc0f21a29be6fb88f627d807470d02d94b.pdf|web_289916b20eaef9df.pdf,8125d1dc0f21a29be6fb88f627d807470d02d94b.pdf|web_289916b20eaef9df.pdf|web_ec4eb8b43a001ba9.pdf|5fdcf120b846746985684fa54250aab44639b30c.pdf,web_b3c7755f9510022a.pdf|web_ae4d50e1f662117f.pdf,web_b3c7755f9510022a.pdf|web_ae4d50e1f662117f.pdf|web_2be6edaf3332034d.pdf|web_bfb031d3e591ca35.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q61/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q61/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q61/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q61/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q61/random_k4
 q62,2305.13186v3.pdf,How many more claims does the Wiki Table datasets have comparing to scientific articles datasets? Please writeh the answer in int format. ,197940,Int,2305.13186v3.pdf,27,web_bfb031d3e591ca35.pdf|70d4f3a5fd97468630edcb4638e7c07b844b85fe.pdf,web_bfb031d3e591ca35.pdf|70d4f3a5fd97468630edcb4638e7c07b844b85fe.pdf|acm_3744614_scientific_factchecking.pdf|arxiv_1909.02164.pdf,web_2fd45ed770cf7563.pdf|a06a34d02fd632c9dc8c629a681d463f3a5bdcb5.pdf,web_2fd45ed770cf7563.pdf|a06a34d02fd632c9dc8c629a681d463f3a5bdcb5.pdf|web_d8f936ecd4367f91.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q62/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q62/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q62/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q62/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q62/random_k4
 q63,2305.13186v3.pdf,How many USD dollars will the author spend if the annotators finish annotating all of claims of the authors' proposed dataset? Please write the answer in float format.,453.25,Float,2305.13186v3.pdf,27,web_bfb031d3e591ca35.pdf|70d4f3a5fd97468630edcb4638e7c07b844b85fe.pdf,web_bfb031d3e591ca35.pdf|70d4f3a5fd97468630edcb4638e7c07b844b85fe.pdf|acm_3744614_scientific_factchecking.pdf|arxiv_1909.02164.pdf,web_2fd45ed770cf7563.pdf|a06a34d02fd632c9dc8c629a681d463f3a5bdcb5.pdf,web_2fd45ed770cf7563.pdf|a06a34d02fd632c9dc8c629a681d463f3a5bdcb5.pdf|web_d8f936ecd4367f91.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q63/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q63/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q63/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q63/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q63/random_k4
-q64,91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf,"A transmitter operating at a frequency of 500 MHz has a rated maximum output power of 10 W. Using the recommended separation distances between portable and mobile RF communications equipment and the M270TF-XXX / M320TF-XXX, calculate the minimum separation distance required for this transmitter. Write down the answer in float format with 2 decimals.",2.92,Float,91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf,40,web_8bce5014fb268e2e.pdf|web_ca82e89a96a0fe65.pdf,web_8bce5014fb268e2e.pdf|web_ca82e89a96a0fe65.pdf|web_f539c13fe18de9dd.pdf|web_1fed1ad9de3ee391.pdf,web_d0752a6a77ae08fd.pdf|web_2be6edaf3332034d.pdf,web_d0752a6a77ae08fd.pdf|web_2be6edaf3332034d.pdf|web_2ccba6013fe960b7.pdf|c726c35b04831158d54cee8acd7202672bb42877.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/random_k4
+q64,91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf,"According to the Technical Specification appendix of the 4K UHD Display User Manual, what is the horizontal pixel pitch (in micrometers) of the M320TF-XX monitor model?",184.5,Float,91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf,40,web_8bce5014fb268e2e.pdf|web_ca82e89a96a0fe65.pdf,web_8bce5014fb268e2e.pdf|web_ca82e89a96a0fe65.pdf|web_f539c13fe18de9dd.pdf|web_1fed1ad9de3ee391.pdf,web_d0752a6a77ae08fd.pdf|web_2be6edaf3332034d.pdf,web_d0752a6a77ae08fd.pdf|web_2be6edaf3332034d.pdf|web_2ccba6013fe960b7.pdf|c726c35b04831158d54cee8acd7202672bb42877.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q64/random_k4
 q65,91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf,What is the second bullet point for safety in healthcare applications for the Winmate-brand monitor?,"""Furthermore, all configurations shall comply with the system standard IEC60601-1-1. Everybody who connects additional equipment to the signal input part or signal output part configures a healthcare system, and is therefore, responsible that the system complies with the requirements of the system standard IEC60601-1-1""",Str,91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf,40,web_8bce5014fb268e2e.pdf|web_ca82e89a96a0fe65.pdf,web_8bce5014fb268e2e.pdf|web_ca82e89a96a0fe65.pdf|web_f539c13fe18de9dd.pdf|web_1fed1ad9de3ee391.pdf,web_d0752a6a77ae08fd.pdf|web_2be6edaf3332034d.pdf,web_d0752a6a77ae08fd.pdf|web_2be6edaf3332034d.pdf|web_2ccba6013fe960b7.pdf|c726c35b04831158d54cee8acd7202672bb42877.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q65/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q65/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q65/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q65/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q65/random_k4
 q66,91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf,"Among the tables in the Appendix of the monitor made by Winmate's user manual, how many columns does the table with the largest layout have?",3,Int,91521110100M_4K_UHD_Display_User_Manual_V1.1.pdf,40,web_8bce5014fb268e2e.pdf|web_ca82e89a96a0fe65.pdf,web_8bce5014fb268e2e.pdf|web_ca82e89a96a0fe65.pdf|web_f539c13fe18de9dd.pdf|web_1fed1ad9de3ee391.pdf,web_d0752a6a77ae08fd.pdf|web_2be6edaf3332034d.pdf,web_d0752a6a77ae08fd.pdf|web_2be6edaf3332034d.pdf|web_2ccba6013fe960b7.pdf|c726c35b04831158d54cee8acd7202672bb42877.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q66/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q66/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q66/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q66/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q66/random_k4
 q67,bdf54dxa.pdf,What is the running time for a Hygiene program wash cycle listed in the manual?,200,Int,bdf54dxa.pdf,54,web_7aa5b865fef4a4b5.pdf|web_c0def62c1d919fa0.pdf,web_7aa5b865fef4a4b5.pdf|web_c0def62c1d919fa0.pdf|web_9a7360a70aa0e44b.pdf|web_d4e82c9f51b32f1d.pdf,web_fe2b87c730235158.pdf|web_7795b10bf9b6a434.pdf,web_fe2b87c730235158.pdf|web_7795b10bf9b6a434.pdf|web_b098652d59ada239.pdf|web_ffc784688ff74004.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q67/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q67/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q67/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q67/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q67/random_k4
@@ -84,7 +84,7 @@ q81,f8d3a162ab9507e021d83dd109118b60.pdf,How many quizzes are there in the Busin
 q82,f8d3a162ab9507e021d83dd109118b60.pdf,"How many quizzes are there in units 4, 5, and 6 of the Business course offered by Chaffey College combined?",2,Int,f8d3a162ab9507e021d83dd109118b60.pdf,17,web_0d779de9374a69ef.pdf|web_b4009fe9cbb76c2a.pdf,web_0d779de9374a69ef.pdf|web_b4009fe9cbb76c2a.pdf|web_33eef0cb801b81ea.pdf|web_bdce50920f7f48d8.pdf,web_ed8636d8f0faddf2.pdf|web_289916b20eaef9df.pdf,web_ed8636d8f0faddf2.pdf|web_289916b20eaef9df.pdf|3becf3cee2ec9db3b4bdae0c4504cf6821a6d8e6.pdf|web_f915df7184680905.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q82/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q82/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q82/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q82/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q82/random_k4
 q83,chapter8-geneticscompatibilitymode-141214140247-conversion-gate02_95.pdf,"Which chapters are included in the slide? Your answer should be formatted as a list like ['Chapter 1', 'Chapter 3'].","['Chapter 8', 'Chapter 9', 'Chapter 10']",List,chapter8-geneticscompatibilitymode-141214140247-conversion-gate02_95.pdf,78,web_07e73e4b2ca8d4e2.pdf|web_d1015c22be42ae71.pdf,web_07e73e4b2ca8d4e2.pdf|web_d1015c22be42ae71.pdf|web_2f9dfa8a6ca32296.pdf|web_5e409f22555df076.pdf,web_4886406d9f17960b.pdf|web_88be333a901981bb.pdf,web_4886406d9f17960b.pdf|web_88be333a901981bb.pdf|web_c0def62c1d919fa0.pdf|176f3f770cd32f4ad62a6354b45cad4e425d6146.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q83/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q83/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q83/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q83/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q83/random_k4
 q84,SnapNTell.pdf,"Compared to ViQuAE, how many more QA pairs does SnapNtell have?",71980,Int,SnapNTell.pdf,20,web_4dcbbf87758697e3.pdf|web_99fd395a0c29c14c.pdf,web_4dcbbf87758697e3.pdf|web_99fd395a0c29c14c.pdf|web_87a18654e37349f2.pdf|web_903701db648fd6ad.pdf,web_88be333a901981bb.pdf|web_18e98cd838edeb2e.pdf,web_88be333a901981bb.pdf|web_18e98cd838edeb2e.pdf|web_3dbc7a33416787aa.pdf|web_ff408b2c8a1714b8.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q84/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q84/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q84/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q84/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q84/random_k4
-q85,SnapNTell.pdf,"In the 3rd Wiki filtering, how many more entities were filtered out compared to the 1st Wiki filtering?",2885,Int,SnapNTell.pdf,20,web_4dcbbf87758697e3.pdf|web_99fd395a0c29c14c.pdf,web_4dcbbf87758697e3.pdf|web_99fd395a0c29c14c.pdf|web_87a18654e37349f2.pdf|web_903701db648fd6ad.pdf,web_88be333a901981bb.pdf|web_18e98cd838edeb2e.pdf,web_88be333a901981bb.pdf|web_18e98cd838edeb2e.pdf|web_3dbc7a33416787aa.pdf|web_ff408b2c8a1714b8.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/random_k4
+q85,SnapNTell.pdf,"According to the entity dataset filtering statistics, how many entities were removed by the 1st Wiki filtering stage (which removes entities without a wiki page)?",4457,Int,SnapNTell.pdf,20,web_4dcbbf87758697e3.pdf|web_99fd395a0c29c14c.pdf,web_4dcbbf87758697e3.pdf|web_99fd395a0c29c14c.pdf|web_87a18654e37349f2.pdf|web_903701db648fd6ad.pdf,web_88be333a901981bb.pdf|web_18e98cd838edeb2e.pdf,web_88be333a901981bb.pdf|web_18e98cd838edeb2e.pdf|web_3dbc7a33416787aa.pdf|web_ff408b2c8a1714b8.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q85/random_k4
 q86,SnapNTell.pdf,How many years after The Chapel Bridge was The Acropolis Museum completed? Return me an integer.,676,Int,SnapNTell.pdf,20,web_4dcbbf87758697e3.pdf|web_99fd395a0c29c14c.pdf,web_4dcbbf87758697e3.pdf|web_99fd395a0c29c14c.pdf|web_87a18654e37349f2.pdf|web_903701db648fd6ad.pdf,web_88be333a901981bb.pdf|web_18e98cd838edeb2e.pdf,web_88be333a901981bb.pdf|web_18e98cd838edeb2e.pdf|web_3dbc7a33416787aa.pdf|web_ff408b2c8a1714b8.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q86/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q86/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q86/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q86/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q86/random_k4
 q87,RAR.pdf,"In the 4-shot classification task, how much higher is the top-5 average accuracy of RAR (LLaVA1.5) compared to the top-1 accuracy?
 
@@ -103,9 +103,9 @@ q98,measuringsuccessonfacebooktwitterlinkedin-160317142140_95.pdf,"What is the S
 q99,measuringsuccessonfacebooktwitterlinkedin-160317142140_95.pdf,"In the two years 2013-2014, in which year is there a lower percentage of users from rural locations on the platform whose measurement tools include Twitter Analytics, Followerwonk, Buffer, and TweetReach?",2013,Int,measuringsuccessonfacebooktwitterlinkedin-160317142140_95.pdf,25,web_1bb55f026f3c5bd5.pdf|web_93a0de1d8da7a0cf.pdf,web_1bb55f026f3c5bd5.pdf|web_93a0de1d8da7a0cf.pdf|web_5da919f043d1386a.pdf|web_55eb3e1bfd6dcec7.pdf,web_830d0aa2115caf34.pdf|1cbbf9ab64afd5e814f75a94d89e5a1f4aa8af62.pdf,web_830d0aa2115caf34.pdf|1cbbf9ab64afd5e814f75a94d89e5a1f4aa8af62.pdf|web_4cf87d006f2e4a65.pdf|70d4f3a5fd97468630edcb4638e7c07b844b85fe.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q99/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q99/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q99/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q99/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q99/random_k4
 q100,StudentSupport_Guidebook.pdf,"According to the PDF source, what is the telephone number of Prime Taxi?",+65 6778-0808,Str,StudentSupport_Guidebook.pdf,44,web_c4b89ce88439ccc8.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_c4b89ce88439ccc8.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|1cbbf9ab64afd5e814f75a94d89e5a1f4aa8af62.pdf|fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf,web_6bb860a08864dd7a.pdf|web_4ae8d7cc08fa2f24.pdf,web_6bb860a08864dd7a.pdf|web_4ae8d7cc08fa2f24.pdf|arxiv_2405_20834.pdf|web_ff408b2c8a1714b8.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q100/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q100/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q100/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q100/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q100/random_k4
 q101,StudentSupport_Guidebook.pdf,"According to the document, how many SG University Holidays from 15 May 2016 to 15 Jan 2017?",5,Int,StudentSupport_Guidebook.pdf,44,web_c4b89ce88439ccc8.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_c4b89ce88439ccc8.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|1cbbf9ab64afd5e814f75a94d89e5a1f4aa8af62.pdf|fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf,web_6bb860a08864dd7a.pdf|web_4ae8d7cc08fa2f24.pdf,web_6bb860a08864dd7a.pdf|web_4ae8d7cc08fa2f24.pdf|arxiv_2405_20834.pdf|web_ff408b2c8a1714b8.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q101/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q101/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q101/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q101/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q101/random_k4
-q102,PS_2018.01.09_STEM_FINAL.pdf,"Among people in STEM jobs,  what is the type of STEP major with the largest gender gap consistent employment in the field that receive the highest pay in the survey of Pew Research Center, January 2018? ",computer science,Str,PS_2018.01.09_STEM_FINAL.pdf,105,web_d4c9f0dc3f6589cc.pdf|sheet_q102_pew_social.pdf,web_d4c9f0dc3f6589cc.pdf|sheet_q102_pew_social.pdf|web_d3545b20cb487e4b.pdf|web_9a644e720d72b089.pdf,web_6c15b8b394b28edf.pdf|web_b873ee670af56beb.pdf,web_6c15b8b394b28edf.pdf|web_b873ee670af56beb.pdf|web_25bb7468c3ed16b0.pdf|web_82328a4273fef79b.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/random_k4
+q102,PS_2018.01.09_STEM_FINAL.pdf,"In the Pew Research Center survey (January 2018), by how many percentage points was the share of women in STEM who said they had experienced workplace discrimination higher than the share of women in non-STEM jobs who said the same?",9,Int,PS_2018.01.09_STEM_FINAL.pdf,105,web_d4c9f0dc3f6589cc.pdf|sheet_q102_pew_social.pdf,web_d4c9f0dc3f6589cc.pdf|sheet_q102_pew_social.pdf|web_d3545b20cb487e4b.pdf|web_9a644e720d72b089.pdf,web_6c15b8b394b28edf.pdf|web_b873ee670af56beb.pdf,web_6c15b8b394b28edf.pdf|web_b873ee670af56beb.pdf|web_25bb7468c3ed16b0.pdf|web_82328a4273fef79b.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q102/random_k4
 q103,e639029d16094ea71d964e2fb953952b.pdf,Since what year has Mr. Kildee been involved with child nutrition?,1954,Int,e639029d16094ea71d964e2fb953952b.pdf,20,web_db15f36044519c81.pdf|web_6c15b8b394b28edf.pdf,web_db15f36044519c81.pdf|web_6c15b8b394b28edf.pdf|web_7b4a136efb01e0d1.pdf|web_caed7cb4152d45e3.pdf,web_0867a04cdab7a273.pdf|web_a39c0d7f5ff0dedb.pdf,web_0867a04cdab7a273.pdf|web_a39c0d7f5ff0dedb.pdf|web_5906edbb3f342646.pdf|web_660e5c500c425b81.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q103/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q103/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q103/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q103/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q103/random_k4
-q104,e639029d16094ea71d964e2fb953952b.pdf,How many committee members from Texas are presented in the conference of improving the quality and efficiency of commodity distribution to federal child nutrition programs?,3,Int,e639029d16094ea71d964e2fb953952b.pdf,20,web_db15f36044519c81.pdf|web_6c15b8b394b28edf.pdf,web_db15f36044519c81.pdf|web_6c15b8b394b28edf.pdf|web_7b4a136efb01e0d1.pdf|web_caed7cb4152d45e3.pdf,web_0867a04cdab7a273.pdf|web_a39c0d7f5ff0dedb.pdf,web_0867a04cdab7a273.pdf|web_a39c0d7f5ff0dedb.pdf|web_5906edbb3f342646.pdf|web_660e5c500c425b81.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/random_k4
+q104,e639029d16094ea71d964e2fb953952b.pdf,"In the House committee whose roster is led by Chairman John A. Boehner of Ohio at the October 7, 2003 hearing, how many of its listed members represent Texas?",3,Int,e639029d16094ea71d964e2fb953952b.pdf,20,web_db15f36044519c81.pdf|web_6c15b8b394b28edf.pdf,web_db15f36044519c81.pdf|web_6c15b8b394b28edf.pdf|web_7b4a136efb01e0d1.pdf|web_caed7cb4152d45e3.pdf,web_0867a04cdab7a273.pdf|web_a39c0d7f5ff0dedb.pdf,web_0867a04cdab7a273.pdf|web_a39c0d7f5ff0dedb.pdf|web_5906edbb3f342646.pdf|web_660e5c500c425b81.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q104/random_k4
 q105,a5879805d70c854ea4361e43a84e3bb2.pdf,what is the email id of the mtre laurent nahmiash ?,lnahmiash@infavocats.com,Str,a5879805d70c854ea4361e43a84e3bb2.pdf,15,web_459cb41b80747aec.pdf|web_9deb1a47e107dc04.pdf,web_459cb41b80747aec.pdf|web_9deb1a47e107dc04.pdf|web_d1b6bfca9d815fc7.pdf|web_5b8076d084922575.pdf,web_757a5d160fcc186b.pdf|web_c4b48f5409e60c35.pdf,web_757a5d160fcc186b.pdf|web_c4b48f5409e60c35.pdf|web_1bb55f026f3c5bd5.pdf|web_88b8d238372b5297.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q105/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q105/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q105/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q105/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q105/random_k4
 q106,a5879805d70c854ea4361e43a84e3bb2.pdf,What exhibit was the plantiff turn 8's representative unaware of?,P-25,Str,a5879805d70c854ea4361e43a84e3bb2.pdf,15,web_459cb41b80747aec.pdf|web_9deb1a47e107dc04.pdf,web_459cb41b80747aec.pdf|web_9deb1a47e107dc04.pdf|web_d1b6bfca9d815fc7.pdf|web_5b8076d084922575.pdf,web_757a5d160fcc186b.pdf|web_c4b48f5409e60c35.pdf,web_757a5d160fcc186b.pdf|web_c4b48f5409e60c35.pdf|web_1bb55f026f3c5bd5.pdf|web_88b8d238372b5297.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q106/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q106/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q106/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q106/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q106/random_k4
 q107,a5879805d70c854ea4361e43a84e3bb2.pdf,"Among the 26 Exhibits from P-1 to P-26, which Exhibits are not mentioned in the document? The answer should be formatted as a list, for example [""P-1"",""P-2""]","['P-15', 'P-16', 'P-17', 'P-18']",List,a5879805d70c854ea4361e43a84e3bb2.pdf,15,web_459cb41b80747aec.pdf|web_9deb1a47e107dc04.pdf,web_459cb41b80747aec.pdf|web_9deb1a47e107dc04.pdf|web_d1b6bfca9d815fc7.pdf|web_5b8076d084922575.pdf,web_757a5d160fcc186b.pdf|web_c4b48f5409e60c35.pdf,web_757a5d160fcc186b.pdf|web_c4b48f5409e60c35.pdf|web_1bb55f026f3c5bd5.pdf|web_88b8d238372b5297.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q107/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q107/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q107/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q107/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q107/random_k4
@@ -116,7 +116,7 @@ q111,bariumswallowpresentation-090810084400-phpapp01_95.pdf,"In the deck's doubl
 q112,competitiveoutcomes-091006065143-phpapp01_95.pdf,What kind of control predicts college football starting status?,"""elephant"" control",Str,competitiveoutcomes-091006065143-phpapp01_95.pdf,48,web_0fc33975f03605a5.pdf|web_28449e05c7149be2.pdf,web_0fc33975f03605a5.pdf|web_28449e05c7149be2.pdf|sheet_q112_mit_strategy_deck.pdf|web_974a02e9fcc23ca9.pdf,web_040fe479bf878a27.pdf|web_9a7360a70aa0e44b.pdf,web_040fe479bf878a27.pdf|web_9a7360a70aa0e44b.pdf|web_009adc9922648ddf.pdf|web_d4c9f0dc3f6589cc.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q112/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q112/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q112/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q112/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q112/random_k4
 q113,2023.findings-emnlp.248.pdf,"For the dataset that uses Prover9 as the symbolic solver, how many examples in this dataset are used for evaluation?",204,Int,2023.findings-emnlp.248.pdf,19,web_39b6cadc4e9ca03e.pdf|web_949ca008fd6aab55.pdf,web_39b6cadc4e9ca03e.pdf|web_949ca008fd6aab55.pdf|web_56044d2e22a63da0.pdf|web_992a098addf13dd9.pdf,fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf|web_7c476895c4da1bbe.pdf,fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf|web_7c476895c4da1bbe.pdf|web_86667079c4fbb42c.pdf|web_0047db342bad4270.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q113/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q113/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q113/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q113/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q113/random_k4
 q114,2023.findings-emnlp.248.pdf,"When using GPT-4 as the base language model, how many datasets does Logic-LM (without self-refinement) outperform the two baseline models in? ",4,Int,2023.findings-emnlp.248.pdf,19,web_39b6cadc4e9ca03e.pdf|web_949ca008fd6aab55.pdf,web_39b6cadc4e9ca03e.pdf|web_949ca008fd6aab55.pdf|web_56044d2e22a63da0.pdf|web_992a098addf13dd9.pdf,fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf|web_7c476895c4da1bbe.pdf,fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf|web_7c476895c4da1bbe.pdf|web_86667079c4fbb42c.pdf|web_0047db342bad4270.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q114/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q114/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q114/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q114/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q114/random_k4
-q115,2023.findings-emnlp.248.pdf,Which dataset used in this paper was proposed in 2022 and all of its logical reasoning problems are multiple-choice questions with 5 options?,ProofWriter,Str,2023.findings-emnlp.248.pdf,19,web_39b6cadc4e9ca03e.pdf|web_949ca008fd6aab55.pdf,web_39b6cadc4e9ca03e.pdf|web_949ca008fd6aab55.pdf|web_56044d2e22a63da0.pdf|web_992a098addf13dd9.pdf,fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf|web_7c476895c4da1bbe.pdf,fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf|web_7c476895c4da1bbe.pdf|web_86667079c4fbb42c.pdf|web_0047db342bad4270.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/random_k4
+q115,2023.findings-emnlp.248.pdf,"In the paper main results table comparing Standard, Chain-of-Thought, and the proposed method across the five logical reasoning datasets, on which dataset does the proposed method (without self-refinement) achieve its highest accuracy when GPT-4 is the underlying LLM?",LogicalDeduction,Str,2023.findings-emnlp.248.pdf,19,web_39b6cadc4e9ca03e.pdf|web_949ca008fd6aab55.pdf,web_39b6cadc4e9ca03e.pdf|web_949ca008fd6aab55.pdf|web_56044d2e22a63da0.pdf|web_992a098addf13dd9.pdf,fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf|web_7c476895c4da1bbe.pdf,fedc1ec70c2286bce325ca21a9bd6c76463635ec.pdf|web_7c476895c4da1bbe.pdf|web_86667079c4fbb42c.pdf|web_0047db342bad4270.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q115/random_k4
 q116,2023.acl-long.386.pdf,"In the fact-checking paper that decomposes each complex claim into a reasoning program of sub-task functions, how many baselines is the proposed method compared against?",7,Int,2023.acl-long.386.pdf,24,acl_2025.acl-long.1270.pdf|web_86f9823f6b6edb2b.pdf,acl_2025.acl-long.1270.pdf|web_86f9823f6b6edb2b.pdf|9613551ec37b9e99f8cb5ff7c2accd7dc38f3ef1.pdf|web_6bc8f0013b78aca4.pdf,web_ab28fbcd92cfea3d.pdf|web_28449e05c7149be2.pdf,web_ab28fbcd92cfea3d.pdf|web_28449e05c7149be2.pdf|web_24ca5dd0f7f78cc6.pdf|web_6bb860a08864dd7a.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q116/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q116/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q116/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q116/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q116/random_k4
 q117,2023.acl-long.386.pdf,Which type of semantic errors has the lowest total percentages in HOVER dataset?,Subtask,Str,2023.acl-long.386.pdf,24,acl_2025.acl-long.1270.pdf|web_86f9823f6b6edb2b.pdf,acl_2025.acl-long.1270.pdf|web_86f9823f6b6edb2b.pdf|9613551ec37b9e99f8cb5ff7c2accd7dc38f3ef1.pdf|web_6bc8f0013b78aca4.pdf,web_ab28fbcd92cfea3d.pdf|web_28449e05c7149be2.pdf,web_ab28fbcd92cfea3d.pdf|web_28449e05c7149be2.pdf|web_24ca5dd0f7f78cc6.pdf|web_6bb860a08864dd7a.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q117/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q117/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q117/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q117/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q117/random_k4
 q118,tacl_a_00660.pdf,"Among the three correction strategies: training-time correction, generation-time correction, and post-hoc correction, which one has the most representative papers in the survey? ",post-hoc correction,Str,tacl_a_00660.pdf,23,web_cfc8463d4b8b98ee.pdf|web_807d71267666ed53.pdf,web_cfc8463d4b8b98ee.pdf|web_807d71267666ed53.pdf|web_1ac94ecc65814303.pdf|web_7bed9c57d39865f1.pdf,web_d8f936ecd4367f91.pdf|web_4cf87d006f2e4a65.pdf,web_d8f936ecd4367f91.pdf|web_4cf87d006f2e4a65.pdf|web_2ed0552c74ec7055.pdf|web_77f281bc58aa8364.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q118/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q118/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q118/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q118/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q118/random_k4
@@ -188,8 +188,8 @@ q182,b3m5kaeqm2w8n4bwcesw-140602121350-phpapp02_95.pdf,Does CFCs causes skin bur
 q183,b3m5kaeqm2w8n4bwcesw-140602121350-phpapp02_95.pdf,Cold is the catalyst for what reaction that also contributes to ozone layer thinning?,the chemical reaction involving CFCs.,Str,b3m5kaeqm2w8n4bwcesw-140602121350-phpapp02_95.pdf,21,web_8198931726741d61.pdf|web_d6ddd446e3ed89c3.pdf,web_8198931726741d61.pdf|web_d6ddd446e3ed89c3.pdf|web_e03b9b74d910087a.pdf|web_3cd8fd0905eadfd3.pdf,web_e9f399fea7556d59.pdf|web_b0b861dfece3f50d.pdf,web_e9f399fea7556d59.pdf|web_b0b861dfece3f50d.pdf|arxiv_2405_20834.pdf|da82ba1659f104383b88b82d6ffaa0cd0c0d177b.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q183/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q183/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q183/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q183/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q183/random_k4
 q184,2310.09158v1.pdf,Authors of this paper analyzed LLM logical reasoning performance and noted issues like hallucination and counterfactual answers. On which two tasks did these authors mainly conduct experiments?,"['event relation extraction', 'deductive reasoning']",List,2310.09158v1.pdf,23,e7ad08848d5d7c5c47673ffe0da06af443643bda.pdf|cac951523d6b7d1bbd014bd8b5e87afd9ecdd318.pdf,e7ad08848d5d7c5c47673ffe0da06af443643bda.pdf|cac951523d6b7d1bbd014bd8b5e87afd9ecdd318.pdf|937154c2b26d9468af7e9903ae4bf4344df4558c.pdf|58a44062a5b5795816b8b85c9c8c3a95c53d18a7.pdf,web_e6e2caff9e6fa899.pdf|web_839cc267cf5fcc63.pdf,web_e6e2caff9e6fa899.pdf|web_839cc267cf5fcc63.pdf|web_7c476895c4da1bbe.pdf|web_784d228de7928d5e.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q184/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q184/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q184/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q184/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q184/random_k4
 q185,2310.09158v1.pdf,Which baseline did the pre-trained Vicuna-13B perform best on MAVEN-ERE?,CoT w. logical constraints ,Str,2310.09158v1.pdf,23,e7ad08848d5d7c5c47673ffe0da06af443643bda.pdf|cac951523d6b7d1bbd014bd8b5e87afd9ecdd318.pdf,e7ad08848d5d7c5c47673ffe0da06af443643bda.pdf|cac951523d6b7d1bbd014bd8b5e87afd9ecdd318.pdf|937154c2b26d9468af7e9903ae4bf4344df4558c.pdf|58a44062a5b5795816b8b85c9c8c3a95c53d18a7.pdf,web_e6e2caff9e6fa899.pdf|web_839cc267cf5fcc63.pdf,web_e6e2caff9e6fa899.pdf|web_839cc267cf5fcc63.pdf|web_7c476895c4da1bbe.pdf|web_784d228de7928d5e.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q185/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q185/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q185/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q185/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q185/random_k4
-q186,2312.04350v3.pdf,What rung is the second ladder of causation refer to?,intervention,Str,2312.04350v3.pdf,28,arxiv_2404.06349.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf,arxiv_2404.06349.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf|sheet_q186_sighan.pdf|88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf,web_6f67e9068ffec005.pdf|web_43df7d1ce2cae9ac.pdf,web_6f67e9068ffec005.pdf|web_43df7d1ce2cae9ac.pdf|web_0047db342bad4270.pdf|web_185a0d92e212a6ed.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/random_k4
-q187,2312.04350v3.pdf,Which model performs the best on Cladder?,GPT-4,Int,2312.04350v3.pdf,28,arxiv_2404.06349.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf,arxiv_2404.06349.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf|sheet_q186_sighan.pdf|88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf,web_6f67e9068ffec005.pdf|web_43df7d1ce2cae9ac.pdf,web_6f67e9068ffec005.pdf|web_43df7d1ce2cae9ac.pdf|web_0047db342bad4270.pdf|web_185a0d92e212a6ed.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/random_k4
+q186,2312.04350v3.pdf,"In the chain-of-thought prompting strategy CAUSALCOT, by how many points does it improve over the performance of vanilla GPT-4 on the benchmark?",8.37,Float,2312.04350v3.pdf,28,arxiv_2404.06349.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf,arxiv_2404.06349.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf|sheet_q186_sighan.pdf|88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf,web_6f67e9068ffec005.pdf|web_43df7d1ce2cae9ac.pdf,web_6f67e9068ffec005.pdf|web_43df7d1ce2cae9ac.pdf|web_0047db342bad4270.pdf|web_185a0d92e212a6ed.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q186/random_k4
+q187,2312.04350v3.pdf,"In the main results table, how many accuracy points higher is the proposed CausalCoT chain-of-thought method than vanilla GPT-4 on Rung 1 (associational) questions? Give the difference to two decimal places.",20.34,Float,2312.04350v3.pdf,28,arxiv_2404.06349.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf,arxiv_2404.06349.pdf|f556c55df76a9558422f00c4171900dc999c892a.pdf|sheet_q186_sighan.pdf|88d4b9368dcf48f2b3e788da0a5f10e87e5a3361.pdf,web_6f67e9068ffec005.pdf|web_43df7d1ce2cae9ac.pdf,web_6f67e9068ffec005.pdf|web_43df7d1ce2cae9ac.pdf|web_0047db342bad4270.pdf|web_185a0d92e212a6ed.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q187/random_k4
 q188,2305.14160v4.pdf,How much does the Anchor Re-weighting method improve the performance of vanilla ICL on average?,16.74,Float,2305.14160v4.pdf,16,a6fe923e92c4e416a233c149cc066999482b5b53.pdf|3215d058d4baa18eec7b5dacbf5c20dffcfeec63.pdf,a6fe923e92c4e416a233c149cc066999482b5b53.pdf|3215d058d4baa18eec7b5dacbf5c20dffcfeec63.pdf|4d0499fe282064db2cfdcc615c5432ad7d329f48.pdf|bc32d5ba50656bbeb36c8608435957789dfe1f4e.pdf,web_ab2f104fb9b1c2fe.pdf|web_6a55c713826db217.pdf,web_ab2f104fb9b1c2fe.pdf|web_6a55c713826db217.pdf|web_0592b193cfe894c0.pdf|web_99fc489414634c09.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q188/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q188/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q188/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q188/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q188/random_k4
 q189,2305.14160v4.pdf,"Using anchor-only context compression, how much can GPT2-XL speed up on AGNews?",2.5x,Str,2305.14160v4.pdf,16,a6fe923e92c4e416a233c149cc066999482b5b53.pdf|3215d058d4baa18eec7b5dacbf5c20dffcfeec63.pdf,a6fe923e92c4e416a233c149cc066999482b5b53.pdf|3215d058d4baa18eec7b5dacbf5c20dffcfeec63.pdf|4d0499fe282064db2cfdcc615c5432ad7d329f48.pdf|bc32d5ba50656bbeb36c8608435957789dfe1f4e.pdf,web_ab2f104fb9b1c2fe.pdf|web_6a55c713826db217.pdf,web_ab2f104fb9b1c2fe.pdf|web_6a55c713826db217.pdf|web_0592b193cfe894c0.pdf|web_99fc489414634c09.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q189/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q189/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q189/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q189/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q189/random_k4
 q190,Guide-for-international-students-web.pdf,What is NTU standing for in this guidebook?,Nottingham Trent University,Str,Guide-for-international-students-web.pdf,76,web_8c27a6695c93326b.pdf|web_ae1b8a373064ff5e.pdf,web_8c27a6695c93326b.pdf|web_ae1b8a373064ff5e.pdf|web_43df7d1ce2cae9ac.pdf|web_77f281bc58aa8364.pdf,web_7795b10bf9b6a434.pdf|web_ca39b8721f6d1255.pdf,web_7795b10bf9b6a434.pdf|web_ca39b8721f6d1255.pdf|web_2c0d954cf4efd312.pdf|web_18e98cd838edeb2e.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q190/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q190/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q190/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q190/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q190/random_k4
@@ -281,6 +281,6 @@ q267,BESTBUY_2023_10K.pdf,"what proportion does Best Buy have female employees i
 q268,BESTBUY_2023_10K.pdf,how many Domestic store does Best Buy have in fiscal 2023?,978,Int,BESTBUY_2023_10K.pdf,75,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf|web_cd0ef8929be1ba45.pdf|web_4eb4afea574790a6.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|web_715e6cf2d4ba3d94.pdf|web_c916dc9b2089aeec.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q268/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q268/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q268/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q268/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q268/random_k4
 q269,BESTBUY_2023_10K.pdf,what is total current assets in FY2023 for Bestbuy? Answer in million.,8802,Int,BESTBUY_2023_10K.pdf,75,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf|web_cd0ef8929be1ba45.pdf|web_4eb4afea574790a6.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|web_715e6cf2d4ba3d94.pdf|web_c916dc9b2089aeec.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q269/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q269/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q269/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q269/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q269/random_k4
 q270,BESTBUY_2023_10K.pdf,what is basic earnings per ordinary share in FY2023 for Bestbuy?,6.31,Float,BESTBUY_2023_10K.pdf,75,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf|web_cd0ef8929be1ba45.pdf|web_4eb4afea574790a6.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|web_715e6cf2d4ba3d94.pdf|web_c916dc9b2089aeec.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q270/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q270/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q270/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q270/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q270/random_k4
-q271,BESTBUY_2023_10K.pdf,what is interest coverage ratio for AMCOR'FY 2020? round your answer to three decimal?,51.286,Float,BESTBUY_2023_10K.pdf,75,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf|web_cd0ef8929be1ba45.pdf|web_4eb4afea574790a6.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|web_715e6cf2d4ba3d94.pdf|web_c916dc9b2089aeec.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/random_k4
+q271,BESTBUY_2023_10K.pdf,"What is the interest coverage ratio (operating income divided by interest expense) for Best Buy for the fiscal year ended January 28, 2023? Round your answer to three decimals.",51.286,Float,BESTBUY_2023_10K.pdf,75,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf|web_cd0ef8929be1ba45.pdf|web_4eb4afea574790a6.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|web_715e6cf2d4ba3d94.pdf|web_c916dc9b2089aeec.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q271/random_k4
 q272,BESTBUY_2023_10K.pdf,"What is effective tax ratio of Best Buy for for the fiscal year ending January 28, 2023?Answer in percentage term, round to one decimal places.",20.7%,Float,BESTBUY_2023_10K.pdf,75,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf|web_cd0ef8929be1ba45.pdf|web_4eb4afea574790a6.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|web_715e6cf2d4ba3d94.pdf|web_c916dc9b2089aeec.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q272/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q272/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q272/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q272/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q272/random_k4
 q273,BESTBUY_2023_10K.pdf,"what is invested capital  of Best Buy for for the fiscal year ending January 28, 2023? Answer in million.",13929 ,Float,BESTBUY_2023_10K.pdf,75,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf,web_88be333a901981bb.pdf|web_839cc267cf5fcc63.pdf|web_cd0ef8929be1ba45.pdf|web_4eb4afea574790a6.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf,web_459cb41b80747aec.pdf|eb30c70c153e1dbc5a0ea9bd5a306750a2b4ebf7.pdf|web_715e6cf2d4ba3d94.pdf|web_c916dc9b2089aeec.pdf,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q273/control_k0,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q273/hard_negative_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q273/hard_negative_k4,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q273/random_k2,https://huggingface.co/datasets/luoojason/jsaj-eval-bundle/resolve/main/q273/random_k4