Datasets:

SinapisAI
/

RobustRouteBench

@@ -1,37 +1,34 @@
-# License and provenance notice
-This compilation uses `license: other`. No single blanket license is asserted
-over all upstream questions, images, or benchmark annotations. The release
-schema, cleaning script, and newly authored metadata follow the repository's
-project license, but upstream content remains governed by its original terms.
-## General text
 Sources include QANTA, MedQA, GPQA, FinQA, PubMedQA, SuperGLUE, FinanceBench,
 PIQA, HellaSwag, SocialIQA, Minerva/Hendrycks mathematics, MMLU-Pro, MathQA,
-TACO, APPS, and the recorded long-context sources. The final dataset card must
-link each redistributed source to its exact version and license. The historical
-`wustar/*` mirrors recorded in the build config are not anonymously resolvable;
-they are provenance identifiers, not evidence of a redistribution license.
 ## Multimodal
-MMR-Bench V2 aggregates multiple upstream benchmarks and does not assert a
-single license over their questions or images. Preserve the source project's
-`LICENSES.md` and original benchmark terms. The formal test sources include
-AI2D, ScienceQA, ScienceQA-IMG, DocVQA, OCR-VQA, and VQAv2/COCO-derived data.
-MMR-4000 additionally includes MMStar, RealWorldQA, SEEDBench2 Plus, SimpleVQA,
-OCRBench, InfoVQA, ChartQAPro, CharXiv, MathVista, MathVision, MathVerse,
-LogicVista, WeMath, MMMU-Pro, BLINK, ERQA, VStarBench, and HallusionBench.
-The inherited MMR notice marks binary redistribution as unverified.
 ## Chinese
 Sources include AlignBench, C-Eval, CIF-Bench, CLUE-C3, CLUE-CHID,
 CLUE-CMRC2018, CLUE-CSL, CLUE-OCNLI, CMB-Clin, ChineseSimpleQA, FinEval,
-FollowBench-zh, GAOKAO-Bench, LawBench, and Chinese LongBench tasks. Preserve
-the original per-dataset licenses and attribution notices.
-Redistribution must be reviewed against each upstream license. When an upstream
-license does not permit redistribution, publish only IDs, derived routing
-labels, embeddings where permitted, and a source resolver.

+# License and source attribution
+This compilation is marked `license: other`. No blanket license is asserted over
+upstream questions, images, or benchmark annotations. Those materials remain
+governed by their original source terms; the code license does not replace them.
+## General Text
 Sources include QANTA, MedQA, GPQA, FinQA, PubMedQA, SuperGLUE, FinanceBench,
 PIQA, HellaSwag, SocialIQA, Minerva/Hendrycks mathematics, MMLU-Pro, MathQA,
+TACO, APPS, and the recorded long-context sources. Source identifiers establish
+attribution; they do not themselves grant redistribution rights.
 ## Multimodal
+MMR-Bench V2 aggregates multiple upstream benchmarks. Its source terms do not
+provide one blanket permission over all questions and image binaries. Training
+sources include MMStar, RealWorldQA, SEEDBench2 Plus, SimpleVQA, OCRBench,
+InfoVQA, ChartQAPro, CharXiv, MathVista, MathVision, MathVerse, LogicVista,
+WeMath, MMMU-Pro, BLINK, ERQA, VStarBench, and HallusionBench.
+Standard-test sources include AI2D, ScienceQA, ScienceQA-IMG, DocVQA, OCR-VQA,
+and VQAv2/COCO-derived data. Preserve the applicable benchmark and image terms
+and attribution when using or redistributing these materials.
 ## Chinese
 Sources include AlignBench, C-Eval, CIF-Bench, CLUE-C3, CLUE-CHID,
 CLUE-CMRC2018, CLUE-CSL, CLUE-OCNLI, CMB-Clin, ChineseSimpleQA, FinEval,
+FollowBench-zh, GAOKAO-Bench, LawBench, and Chinese LongBench tasks.
+Research, commercial use, and redistribution must comply with each applicable
+upstream license. No additional permission over third-party content is implied
+by this dataset card. Source links and citations are provided in the Dataset Card.

README.md CHANGED Viewed

@@ -32,231 +32,167 @@ configs:
 ![SinapisAI RobustRouteBench: a benchmark for robust model selection across General Text, Multimodal, and Chinese.](assets/robustroutebench-banner.png)
-RobustRouteBench is a unified public benchmark for general-text, multimodal,
-and Chinese-language model routing.
-```python
-from datasets import load_dataset
-text = load_dataset("SinapisAI/RobustRouteBench", "text")
-multimodal = load_dataset("SinapisAI/RobustRouteBench", "multimodal")
-chinese = load_dataset("SinapisAI/RobustRouteBench", "chinese")
-```
-Each scenario exposes separate `train`, `validation`, and `test` splits. The
-500-row validation split is a convenience overlay containing 400 standard and
-100 robust queries. Those rows intentionally remain in the complete test split,
-so validation and test overlap by design.
-The test split contains both standard and robustness evaluation rows. Use
-`evaluation_type` (`standard` or `robust`) and `variant_type` to report slices.
-Each row is query-centric. `performance`, `cost_cny`, and `latency_ms` are
-aligned to `model_ids`. All costs use CNY; historical USD proxies use the frozen
-benchmark conversion `1 USD = 7.2 CNY`.
-Source outcomes that were not successful have performance zero. No source
-diagnostic payloads or raw model responses are published. General Text and Chinese
-use BGE-M3 features. Multimodal uses frozen
-[CLIP ViT-B/32](https://huggingface.co/openai/clip-vit-base-patch32) features
-([ICML paper](https://proceedings.mlr.press/v139/radford21a.html)), checkpoint
-`3d74acf9a28c67741b2f4f2ea7635f0aaf6f0268`. Full prompt tokens are split into
-75-token payload chunks plus BOS/EOS. Normalize each 512-dimensional text chunk
-vector and pool with payload-token-count weights. Encode every referenced RGB
-image using official 224-pixel resizing/center-cropping, normalize each vector,
-and pool images equally. Normalize the two pooled modality vectors, concatenate,
-and divide by √2. This yields 1,024-dimensional **late fusion**, not Qwen joint
-encoding. All methods/seeds use the same features. Missing images are errors;
-labels and reference answers never enter the encoder.
-The corrected release restores exactly **20,167** successful General Text scores
-whose model, request, and answer match the historical nonzero record, including
-522 cells in the validation overlay. Known failed requests retain zero scores.
-Other legacy zeros without a verified scoring source remain unchanged. This is a
-provenance-confirmed repair, **not independent regrading of every legacy outcome**.
-The repository is self-contained. Original multimodal images and accompanying
-metadata are grouped under [media/multimodal/](media/multimodal/README.md):
-`train.zip` (training), `standard_test.zip` (standard test), and `robust_test.zip`
-(robustness test). Media downloads are optional for router training and prediction
-replay with the published CLIP features; they are needed to inspect images,
-rebuild features, or re-evaluate visual models. No separate dataset repository is
-required. See `LICENSES.md`
-for upstream attribution and terms. This compilation uses `license: other`; no
-blanket license is asserted over all upstream questions or images.
-## Evaluation guide
-Companion to **Toward Collective Model Intelligence: Characterizing and Coordinating Models**, Section 6.3 (October 2026).
-This card documents the RobustRouteBench experiment in Section 6.3 of the SinapisAI technical report, including the protocol and complete results previously collected in its RobustRouteBench appendix. The scored benchmark evaluates **offline request-level model selection** under a fixed candidate pool. It covers semantic rewrites, language changes, context noise, and multimodal perturbations. Model-ecosystem events and service recovery are complementary **extension protocols**, with no numerical event-replay or live-service results in the reported snapshot.
-**Mio selector** means the request-level selection component trained separately on each scenario's training split, with hyperparameters and operating points selected using development data. These results do not evaluate the hosted Mio 1.0 Lite API or the end-to-end SinapisAI platform. Mio's internal training and inference implementation is not included in this dataset release.
-The reported experiment covers the fixed-pool, single-request setting with \\(T=1\\) under monetary budgets. Model-side evidence is learned from benchmark training observations; the Section 3 capability-profile construction pipeline is not jointly evaluated here. The results do not isolate profile encoding, pairwise supervision, or near-tie selection, and do not evaluate pool organization, multi-step handoff, or the Section 4.2 rank-fusion rule.
-- [Evaluation populations and development exclusion](#evaluation-populations-and-development-exclusion)
-- [Benchmark composition and transformations](#benchmark-composition-and-transformations)
-- [Compared implementations](#compared-implementations)
-- [Metrics and reporting conventions](#metrics-and-reporting-conventions)
-- [Complete reported results](#complete-reported-results)
-- [Attribution and uncertainty protocol](#attribution-and-uncertainty-protocol)
-- [Model-ecosystem and service extensions](#model-ecosystem-and-service-extensions)
-- [Reproducibility and release boundaries](#reproducibility-and-release-boundaries)
-## Evaluation populations and development exclusion
-**The published `test` split is the complete evaluation inventory, not the report's final holdout.** Its overlap with the 500-row `validation` overlay is intentional. To reproduce the report's final evaluation, remove every source group represented in validation, including linked base requests and variants, before calculating any test metric. Removing only rows with `is_validation=True` is insufficient for the paired text scenarios.
-| Scenario / config | Fixed candidate models | Train | Published standard | Final standard | Published robust | Final robust | Unique scenario instances |
-| --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
-| General Text / `text` | 63 | 3,000 | 4,188 | 3,688 | 3,000 | 2,894 | 10,188 |
-| Multimodal / `multimodal` | 45 | 4,000 | 5,000 | 4,600 | 2,000 | 1,900 | 11,000 |
-| Chinese / `chinese` | 10 | 7,590 | 4,400 | 3,900 | 2,100 | 2,000 | 14,090 |
-| Total | Scenario-specific | 14,590 | 13,588 | 12,188 | 7,100 | 6,794 | 35,278 |
-Validation contains 400 standard and 100 robust requests per scenario and is already counted in the published evaluation inventory. Group exclusion removes 606 General Text, 500 Multimodal, and 600 Chinese rows, leaving **18,982 final-test requests** (6,582 / 6,500 / 5,900). The resulting development-to-test source-group overlap is zero. Document- and repository-level identities are not consistently available, so this is a base-query/source-sample separation claim, not a stronger document-level independence claim. Development used a public subset of the same benchmark release; this is not an independently administered hidden benchmark.
-Training groups are disjoint from development and final-test groups under the same base-query/source-sample identifiers. Benchmark sources and task domains can still be shared across these roles; the group separation does not establish cross-benchmark or cross-domain generalization.
-The grouping rule in the frozen result archive is `base_query_id` when present, otherwise `query_id`, for General Text and Chinese; Multimodal uses `source_sample_id`. The following read-only filtering example operates on the loaded datasets above and does not change any published split:
 ```python
 from collections import Counter
-expected = {
-    "text": {"standard": 3688, "robust": 2894},
-    "multimodal": {"standard": 4600, "robust": 1900},
-    "chinese": {"standard": 3900, "robust": 2000},
-}
-def source_group(row, scenario):
-    if scenario == "multimodal":
-        key = row["source_sample_id"]
-        if key is None or key == "":
-            raise ValueError("Missing multimodal source group")
-    else:
-        key = row["base_query_id"] or row["query_id"]
-    return (scenario, key)
-final_tests = {}
-for scenario, ds in [("text", text), ("multimodal", multimodal), ("chinese", chinese)]:
-    dev_groups = {source_group(row, scenario) for row in ds["validation"]}
-    final = ds["test"].filter(
-        lambda row: source_group(row, scenario) not in dev_groups
-    )
-    assert not dev_groups.intersection(source_group(row, scenario) for row in final)
-    assert Counter(final["evaluation_type"]) == expected[scenario]
-    final_tests[scenario] = final
-```
-Each scenario has its own fixed action space, shared by its training, standard, and robust populations. The pools may overlap in model family or version; **63, 45, and 10 are not one global pool**. Candidate identities, train/development roles, final requests, generation settings, and outcome matrices are held fixed across methods within a scenario. Test utilities, references, and realized costs are evaluation labels and must remain unavailable to the selector at decision time.
-## Benchmark composition and transformations
-### General Text
-| Domain | Train | Published standard | Published robust |
-| --- | ---: | ---: | ---: |
-| Code | 1,386 | 859 | 268 |
-| General Reasoning | 435 | 649 | 474 |
-| Knowledge QA | 388 | 616 | 506 |
-| Long Context | 0 | 1,000 | 1,000 |
-| Mathematical Reasoning | 438 | 574 | 410 |
-| Reading Understanding | 353 | 490 | 342 |
-The 3,000 robust instances derive from 2,907 unique base queries retained in standard test. All 1,000 Long Context queries serve as robustness bases, and none appear in training. The language-variation set contains 999 instances with Chinese characters. Long-context noise inserts one to three background notes whose removal exactly recovers the original request.
-### Multimodal
-MMR-4000 training examples are sampled from 18 [MMR-Bench-V2][ma2026mmrbench] sub-benchmarks. VisionQA-5000 standard evaluation has the following composition:
-| Capability group | Source benchmark | Published standard instances |
-| --- | --- | ---: |
-| General VQA | [VQAv2][goyal2017vqav2] | 1,500 |
-| Science / Diagram | [AI2D][kembhavi2016ai2d] | 676 |
-| Science / Diagram | [ScienceQA][lu2022scienceqa] | 667 |
-| Science / Diagram | [ScienceQA-IMG][lu2022scienceqa] | 657 |
-| Document / OCR | [DocVQA][mathew2021docvqa] | 183 |
-| Document / OCR | [OCR-VQA][mishra2019ocrvqa] | 1,317 |
-| Total | | 5,000 |
-A multimodal request is an ordered sequence of text and visual segments and may contain multiple images and answer options. The 2,000 robust items are sampled **independently** of VisionQA-5000, using the same 45-model action space. Standard versus robust differences are distribution-level comparisons, not per-request clean-versus-perturbed losses.
-### Chinese
-| Coverage | Source benchmark |
 | --- | --- |
-| Alignment | [AlignBench][liu2023alignbench] |
-| Knowledge and examinations | [C-Eval][huang2023ceval] |
-| Instruction following | [CIF-Bench][li2024cifbench] |
-| Language, NLI, and reading | [CLUE benchmark family][xu2020clue] |
-| Clinical medicine | [CMB-Clin from CMB][wang2023cmb] |
-| Factual knowledge | [Chinese SimpleQA][he2025chinesesimpleqa] |
-| Financial knowledge | [FinEval][guo2023fineval] |
-| Constraint following | [FollowBench-zh][jiang2024followbench] |
-| Standardized examinations | [GAOKAO-Bench][zhang2023gaokao] |
-| Legal knowledge | [LawBench][fei2023lawbench] |
-| Long-context tasks | [Chinese subsets of LongBench][bai2023longbench] |
-Each source retains its task and evaluator semantics; the utility is not assumed to be the same accuracy measure across all tasks. The published `performance` arrays are the scored outcomes, with unsuccessful source outcomes set to zero as specified above.
-### Robustness groups and pairing
-| Scenario | Robustness type | Published instances | Final-test instances | Relation to standard population |
-| --- | --- | ---: | ---: | --- |
-| General Text | Semantic rewrite | 1,000 | 963 | Paired base retained |
-| General Text | Language variation | 1,000 | 964 | Paired base retained |
-| General Text | Long-context noise | 1,000 | 967 | Paired base retained |
-| Multimodal | Text rewrite | 500 | 475 | Independently sampled |
-| Multimodal | Context noise | 500 | 475 | Independently sampled |
-| Multimodal | Image variation | 500 | 475 | Independently sampled |
-| Multimodal | Combined perturbation | 500 | 475 | Independently sampled |
-| Chinese | Semantic rewrite | 700 | 667 | Paired, disjoint bases |
-| Chinese | English translation | 700 | 666 | Paired, disjoint bases |
-| Chinese | Prefix noise | 700 | 667 | Paired, disjoint bases |
-Transformations must preserve meaning, necessary evidence, the answer contract, and scorer validity. Image variation assigns 100 instances each to JPEG degradation, controlled cutout, Gaussian noise, Gaussian blur, and contrast reduction. Combined multimodal perturbations change text, context, and image together. The three Chinese types use disjoint base requests, all retained in standard test. Even for the paired text scenarios, the overall standard population contains requests outside the robust base set, so subtracting overall standard and robust metrics is not a paired transformation effect.
-## Compared implementations
-The report ranks **13 distinct implementations**: Mio selector and 12 baselines evaluated through [ORBIT][lai2026orbit]'s common interfaces. All use the same scenario-specific candidate pool and available training/development evidence.
-| Reported implementation | Method source |
 | --- | --- |
 | Avengers | [The Avengers][zhang2025avengers] |
 | Avengers-Pro | [Performance-Efficiency Optimized Routing][zhang2025avengerspro] |
-| Causal RM-Softmax | [Causal LLM Routing, RM-Softmax][tsiourvas2025causalrouting] |
 | Eagle | [Eagle][zhao2024eagle] |
 | EmbedLLM-RouteJudge | [EmbedLLM][zhuang2024embedllm] |
-| Mio selector | Request-oriented selection, technical report Section 4.2; scenario-specific offline training |
 | GraphRouter | [GraphRouter][feng2024graphrouter] |
 | HybridLLM | [Hybrid LLM][ding2024hybridllm] |
-| kNN | [TensorOpera][stripelis2024tensoropera] / [ORBIT][lai2026orbit] baseline |
-| MLP | [TensorOpera][stripelis2024tensoropera] / [ORBIT][lai2026orbit] baseline |
-| MIRT | [IRT-Router, MIRT variant][song2025irtrouter] |
-| RouteLLM-MF | [RouteLLM, matrix-factorization variant][ong2024routellm] |
-| SVM Router | [TensorOpera][stripelis2024tensoropera] / [ORBIT][lai2026orbit] baseline |
-The implementation audit found five groups of additional labels with identical adapters/configurations and identical `model_ids`, performance-score, cost-score, latency-score, and selection-policy arrays across every evaluated scenario and seed:
-| Ranked implementation | Additional implementation aliases in this experiment |
-| --- | --- |
-| Causal RM-Softmax | RM-Classification, RM-Interval ([source][tsiourvas2025causalrouting]) |
-| Eagle | RouteLLM-SW Ranking ([RouteLLM][ong2024routellm]) |
-| EmbedLLM-RouteJudge | Model-SAT ([Capability Instruction Tuning][zhang2025modelsat]), [RouterDC][chen2024routerdc] |
-| MLP | [OmniRouter][mei2025omnirouter], RouteLLM-BERT ([RouteLLM][ong2024routellm]) |
-| MIRT | NIRT ([IRT-Router][song2025irtrouter]) |
-These are aliases **of the fitted implementations in this experiment**, not a claim that the original research methods are equivalent. They receive no extra rows or rank weight and are not independent reproductions of every cited method. Each adapter must declare its input features, fitted parameters, predicted scores, selection policy, and deviations from the source method.
-## Metrics and reporting conventions
-For each method and reporting group, evaluate 100 evenly spaced nominal monetary budgets from the cheapest to the most expensive fixed-model mean cost on that group's held-out requests. These endpoints use evaluation-population cost aggregates, not training/development costs, and are shared across methods. They set the evaluator's reporting grid without supplying request-specific held-out outcomes to selectors. Write the common grid as \\(\mathcal G=\{C^{(1)},\ldots,C^{(K)}\}\\) with \\(K=100\\). Predicted cost determines candidate feasibility; predicted utility or preference selects the model. Held-out outcomes then determine mean utility \\(Q_\pi(C)\\) and realized mean cost \\(\bar c_\pi(C)\\). A nominal budget is not the measured expenditure, and predicted-cost feasibility does not guarantee a realized-cost bound.
-For this offline monetary-budget sweep, a request with no candidate satisfying the predicted cost budget falls back to the model with the lowest predicted cost; its measured quality and cost remain in the reported point. This fallback can exceed the nominal budget even in predicted cost. It is an evaluation convention, distinct from the Section 4.2 deployment rule that reports unmet constraints when no feasible candidate exists.
-Let \\(Q_{\mathrm{strongest}}\\) and \\(\bar c_{\mathrm{strongest}}\\) be the quality and mean cost of the strongest fixed model on the same split and candidate pool. The primary metrics are:
 $$
 \begin{aligned}
@@ -265,73 +201,54 @@ Q_{\mathrm{peak}}(\pi)&=\max_{C\in\mathcal G}Q_\pi(C),\\
 \end{aligned}
 $$
-Cost@90, Cost@95, and Cost@100 use \\(\tau=0.90,0.95,1.00\\), respectively. An unattained target is **N/R**; no extrapolation supplies a missing operating point. For a reached target:
 $$
 \operatorname{Saving}_{\pi}\mathbin{@}\tau=1-\frac{\bar{c}_{\pi}\mathbin{@}\tau}{\bar c_{\mathrm{strongest}}}.
 $$
-Positive savings mean lower cost than always selecting the strongest fixed model; negative savings mean higher cost. Savings are also N/R when the target is unreached.
-For nAUC, first form the cumulative maximum of quality in increasing **nominal-budget** order:
-$$
-Q_\pi^{\mathrm{env}}(C^{(g)})=\max_{1\leq h\leq g}Q_\pi(C^{(h)}).
 $$
-Then integrate by the trapezoidal rule and divide by the full nominal-budget interval:
-$$
-\operatorname{nAUC}(\pi)=\frac{1}{C^{(K)}-C^{(1)}}\sum_{g=1}^{K-1}
 \frac{Q_\pi^{\mathrm{env}}(C^{(g)})+Q_\pi^{\mathrm{env}}(C^{(g+1)})}{2}
 \bigl(C^{(g+1)}-C^{(g)}\bigr).
 $$
-This guarantees \\(\operatorname{nAUC}(\pi)\leq Q_{\mathrm{peak}}(\pi)\\). Fixed-model references, artificial endpoints, and realized-cost completion points never enter the router curve. The report's quality-cost plot instead uses realized mean cost on a logarithmic x-axis; its geometric area is **not** nAUC. For that visualization, duplicate realized-cost points retain the highest quality, dominated points are removed, and the frontier is not extrapolated. Baselines are plotted at their least-cost peak-quality points, with the strongest baseline by nAUC highlighted separately on each split.
-A diagnostic oracle selects the best observed model independently for each request:
-$$
-Q_{\mathrm{oracle}}=\frac{1}{|\mathcal X|}\sum_{x\in\mathcal X}\max_{f_i\in\mathcal C}y_i(x),
-$$
-where \\(\mathcal X\\) is the reporting population, \\(\mathcal C\\) is its candidate pool, and \\(y_i(x)\\) is the held-out utility. The oracle supplies no information to a selector. An evaluation record should retain the raw AUC, nominal interval endpoints, and strongest/cheapest fixed-model and oracle diagnostics so normalization can be reconstructed.
-**Currency.** Published `cost_cny` arrays always use CNY. The frozen report tables below retain their original units: USD for General Text and Multimodal, CNY for Chinese. Divide the first two scenarios' `cost_cny` values by **7.2** to recover those reporting units; convert budget values consistently. This is a frozen benchmark constant, not a current exchange rate or the pricing convention of the report's separate Mio 1.0 Lite experiment. A uniform currency rescaling of costs and budgets leaves nAUC unchanged. Multimodal costs are frozen USD proxies, not live service quotes.
-**Evaluator compatibility.** The accompanying Model Intelligence Toolkit evaluator computes a realized-cost Pareto AUC with a cheapest-fixed-model reference and endpoint completion. That quantity is not the report's nominal-budget nAUC above; its threshold costs can also include the inserted reference. To match this report, use actual router budget points only, align repeated runs on their common nominal-budget grid, average the utility and realized-cost curves pointwise, and then compute peak quality, nAUC, and target costs from that aggregate curve. A field named `normalized_auc` in another evaluator is not sufficient to establish metric compatibility.
-**Repetitions and uncertainty.** Multimodal uses three fitted runs for every implementation, with seeds 20260917, 20260918, and 20260919, 4,000 training requests, 45 candidates, shared CLIP text/image fusion features, and the same evaluation requests and 100-point grid for each reporting population. Fitting seeds vary; evaluation populations, source-group exclusions, and candidate identities are fixed. The reported curve is the pointwise mean of those three aligned run curves, formed before the cumulative maximum or any curve statistic is computed. Quality entries show the aggregate-curve statistic ± its exact nonparametric run-bootstrap standard error, obtained from all \\(3^3\\) ordered resamples of the run curves. General Text and Chinese each use one fitted run per method (seed 20260917). These errors describe fitting variation, not request-sampling uncertainty. Target costs are computed from the aggregate curve; individual-run reach counts are retained only as diagnostics. Current tables contain no paired request-bootstrap intervals or statistical-significance tests.
-**Aggregation and latency.** Macro nAUC is the unweighted mean of the six standard/pooled-robust split values, not a pooled curve over different candidate pools, units, or budget ranges. Ties share rank. Latency is auxiliary: realized latency at each budget and at peak, total router inference time, and mean router time per request are useful diagnostics, but they neither determine the primary ranking nor impose a hard latency constraint in this experiment.
-Robust split metrics are computed directly from the pooled robust request population. Each robustness type is evaluated separately with cost endpoints from that type's requests. Consequently, both the nominal grid and the cumulative-maximum envelope can differ, and averaging type-level nAUC values does not generally recover pooled Robust nAUC. The Multimodal robustness types each contain 475 final-test requests; equal group sizes do not change this aggregation order.
-The wider reporting protocol calls for standard, each robust type, robust pooled, each task domain, and standard-plus-robust groups. A combined Multimodal statistic must be labeled an **unpaired aggregate**. A weakest-group statistic takes the minimum over groups declared before final testing and names the attaining group; an empty group is unavailable, not zero. The tables below contain the released primary and Mio variation-level results, not every diagnostic proposed by that protocol.
-## Complete reported results
-The corrected result snapshot is dated **2026-10-15**. Mio selector has macro
-nAUC **0.6360**, followed by MLP at **0.6183**. Mio leads pooled General Text
-Robust and both Multimodal splits, but not General Text Standard or either
-Chinese split. It also does not lead every perturbation type: language-variation
-nAUC ranks second and Multimodal combined-perturbation nAUC ranks third.
-Public baselines use fixed configurations; Mio uses designated development data
-for configuration selection, not a matched hyperparameter-search budget.
-![Quality-cost frontiers for General Text, Multimodal, and Chinese on standard and robust splits, comparing Mio selector with the 12 distinct baselines.](assets/quality-cost-results.png)
-*Green: Mio selector frontier and peak. Hollow circles: baseline peaks. Blue diamond: best baseline by nAUC. Multimodal points summarize the pointwise-mean curve over three fitted runs. The log x-axis shows realized cost (USD for General Text and Multimodal; CNY for Chinese); the plotted area is not nominal-budget nAUC.*
-The following tables are generated from the corrected, public-data evaluations.
-All 12 baselines were refitted per scenario; Text Mio was refitted/recalibrated,
-Multimodal Mio was refitted with CLIP features, and unchanged Chinese Mio
-predictions were reevaluated. Costs are per request in the named currency.
-**N/R** means the aggregate curve does not reach a threshold; `(r/3)` records
-individual-run reach counts. Primary-table rank is by nAUC.
 <!-- BEGIN GENERATED RESULTS -->
 ### Macro nAUC
 | Rank | Implementation | Macro nAUC |
@@ -486,75 +403,28 @@ Ranks are among 13 distinct implementations; Multimodal metrics use the pointwis
 | Chinese | Semantic rewrite | 0.6156 | 5/13 | 0.6086 | 4/13 |
 | Chinese | English translation | 0.5893 | 7/13 | 0.5802 | 5/13 |
 | Chinese | Prefix noise | 0.6375 | 7/13 | 0.6287 | 5/13 |
 <!-- END GENERATED RESULTS -->
-## Attribution and uncertainty protocol
-This section specifies analyses supported by the benchmark design. The current primary tables do **not** report the paired decomposition, bootstrap confidence intervals, or a complete weakest-group analysis.
-For a complete routed execution, let \\(y_\pi(x)\\) be utility and let nonnegative analysis weights \\(w(x)\\) sum to one over a declared population \\(\mathcal X\\). A paired policy comparison uses:
-$$
-\bar y_\pi=\sum_{x\in\mathcal X}w(x)y_\pi(x),\qquad
-\Delta_y=\sum_{x\in\mathcal X}w(x)\bigl[y_{\pi_1}(x)-y_{\pi_2}(x)\bigr].
-$$
-Weights encode the source-group and within-group estimand and must be fixed independently of final outcomes. A cluster bootstrap resamples source groups within task strata while keeping variants and competing policies' outcomes together. Repeated training seeds quantify fitting variation separately. Methods, thresholds, group definitions, and interval procedures should be frozen before final evaluation.
-For paired General Text or Chinese requests, let \\(\mathsf T\\) be a validated transformation and \\(\pi(x)\\) the selected model index with context, service snapshot, and allowance fixed. Total paired change decomposes as:
-$$
-\begin{aligned}
-\Delta_{\mathrm{model}}(x)&=y_{\pi(x)}(\mathsf T(x))-y_{\pi(x)}(x),\\
-\Delta_{\mathrm{selection}}(x)&=y_{\pi(\mathsf T(x))}(\mathsf T(x))-y_{\pi(x)}(\mathsf T(x)).
-\end{aligned}
-$$
-The first term measures changed performance of the originally selected model; the second measures the additional effect of changing the choice. Their sum is the total paired change. A changed selection is therefore not itself a failure. The identity requires matched outcomes and does not apply to the independently sampled Multimodal sets. Attributing candidate-organization effects requires the staged ecosystem controls below.
-## Model-ecosystem and service extensions
-These are controlled experiment specifications, **not completed numerical evaluations**. Each starts from a frozen snapshot and replays an ordered event stream, recording both effective time and information-visibility time. Conditions receive the same events and may use only evidence visible before the next decision. Preparation/update costs remain separate from subsequent execution costs.
-| Condition | Permitted update | Intended attribution |
-| --- | --- | --- |
-| Frozen | None | Exposure of the original system to the event |
-| Capability refresh | New measurements or reconstructed capability profiles | Benefit of updated capability evidence |
-| Candidate repair | Capability refresh plus candidate-pool reconstruction | Additional effect of changing the retained action set |
-| Selector adaptation | Prior updates plus fitting or calibration | Additional effect of adapting request-level choices |
-| Execution recovery | Endpoint rescheduling; explicit selector reselection stated separately | Restored service versus a changed model choice |
-The ecosystem stream covers model arrival, withdrawal, version replacement, profile refresh, and evaluation-scope expansion. It tests capability characterization, candidate organization, and selector adaptation cumulatively.
-The service stream controls load, prices, rate limits, endpoint interruption/withdrawal, SLOs, and budgets. Compare frozen endpoint assignment, same-model rescheduling with bounded recovery, and a condition explicitly permitting a new Mio selector decision. Over a common post-event horizon, report task utility, expenditure, SLO attainment, throughput, tail latency, retries, endpoint switches, reselections, and recovery time. Unrecovered episodes retain failure or censoring status. Stored outcome matrices can replay selection alternatives; they cannot establish live queueing, failover delay, cancellation, or readiness.
-## Reproducibility and release boundaries
-Use the current **v1.1.2-20261015** release for reproducible experiments.
-This packaging update standardizes media archive filenames and their row-level references; request contents, scores, embeddings, predictions, and image payloads are unchanged.
-Resolve and pin its Git revision through `revision=` when loading data for a fixed
-experiment. `manifest.json` records current file checksums; `release-audit.json`
-records structural/range/group/feature validation and the limits of score verification.
-Split inventories and image payloads are unchanged; the provenance-confirmed Text score
-repair and CLIP Multimodal features remain in effect.
-The numerical source is `report/aggregate-results.json`, schema
-`routestressbench-report-v3`, with the same content used to generate the report
-tables and figure. `report/reproduction-manifest.json` identifies every ID-aligned,
-SHA256-checked per-run prediction file. Public baseline configurations and fitting
-code are in the [companion repository](https://github.com/SinapisAI/collective-model-intelligence).
-The companion command `python -m examples.reproduce_report --mode replay --output runs/replay`
-recomputes report metrics from public data and prediction arrays. `--mode fit`
-re-trains the 12 public baselines; Mio is explicitly replay-only. Library/hardware
-differences can affect fitted neural outputs. Mio's internal training implementation,
-raw responses, private source diagnostics, and completed live-service traces are
-not supplied. Offline score replay does not reproduce live-service experiments.
-For a fully auditable new run, retain split/pair identifiers, source groups, the development-group exclusion list, model/endpoint versions and candidate membership, transformations, seeds, generation/scoring settings, price tables, response status and available latency fields, method configurations, the ORBIT/toolkit commit, source-method deviations, task-level outputs, and prediction-array hashes for alias detection. Dynamic studies additionally need event ordering, visibility times, update costs, and execution traces. This is the required experimental record, not a claim that every listed item is present in the published dataset manifest.
-Upstream question and media attribution and terms remain in [`LICENSES.md`](https://huggingface.co/datasets/SinapisAI/RobustRouteBench/blob/main/LICENSES.md).
 [bai2023longbench]: https://arxiv.org/abs/2308.14508
 [chen2024routerdc]: https://proceedings.neurips.cc/paper_files/paper/2024/hash/7a641b8ec86162fc875fb9f6456a542f-Abstract-Conference.html
@@ -587,3 +457,4 @@ Upstream question and media attribution and terms remain in [`LICENSES.md`](http
 [zhang2025modelsat]: https://ojs.aaai.org/index.php/AAAI/article/view/34790
 [zhao2024eagle]: https://arxiv.org/abs/2409.15518
 [zhuang2024embedllm]: https://proceedings.iclr.cc/paper_files/paper/2025/hash/bf5e4b85d203481d6e37bd32d9600162-Abstract-Conference.html

 ![SinapisAI RobustRouteBench: a benchmark for robust model selection across General Text, Multimodal, and Chinese.](assets/robustroutebench-banner.png)
+**RobustRouteBench** evaluates request-level model selection under quality and cost
+trade-offs across General Text, Multimodal, and Chinese. It accompanies Section 6.3
+of **Toward Collective Model Intelligence: Characterizing and Coordinating Models**.
+[Technical report and code](https://github.com/SinapisAI/collective-model-intelligence) ·
+[Data fields](normalized-schema.json) · [Multimodal images](media/multimodal/README.md) ·
+[Source attribution and terms](LICENSES.md)
+## Dataset overview
+| Scenario / config | Models | Train | Validation | Standard test | Robust test |
+| --- | ---: | ---: | ---: | ---: | ---: |
+| General Text / `text` | 63 | 3,000 | 500 | 3,688 | 2,894 |
+| Multimodal / `multimodal` | 45 | 4,000 | 500 | 4,600 | 1,900 |
+| Chinese / `chinese` | 10 | 7,590 | 500 | 3,900 | 2,000 |
+| Total | Scenario-specific | 14,590 | 1,500 | 12,188 | 6,794 |
+The dataset contains **35,072 request instances**, including **18,982 test
+requests**. Each scenario has its own candidate pool. The test populations above
+are exactly those used in the technical report.
+`train` is used for fitting selectors, `validation` for configuration selection,
+and `test` for reporting results. Each validation split contains 400 standard and
+100 robust requests. The supplied test split is ready for evaluation: **use all
+its rows directly**. Standard and robust results are slices of that same test
+split, identified by `evaluation_type`.
+## Quick start
 ```python
 from collections import Counter
+from datasets import load_dataset
+ds = load_dataset("SinapisAI/RobustRouteBench", "text")
+print({split: len(rows) for split, rows in ds.items()})
+# {'train': 3000, 'validation': 500, 'test': 6582}
+print(Counter(ds["test"]["evaluation_type"]))
+# Counter({'standard': 3688, 'robust': 2894})
+standard = ds["test"].filter(lambda row: row["evaluation_type"] == "standard")
+robust = ds["test"].filter(lambda row: row["evaluation_type"] == "robust")
+```
+Use `"multimodal"` or `"chinese"` for the other scenarios. Their test sizes are
+6,500 and 5,900, respectively.
+## Data fields and units
+Each row contains one request and aligned outcomes for its candidate models.
+| Field | Meaning |
 | --- | --- |
+| `query_id` | Request identifier, shared with the prediction and media files. |
+| `scenario`, `published_split` | Scenario and split membership (`train`, `validation`, or `test`). |
+| `query_text`, `input_json` | Full request text and structured input. |
+| `evaluation_type`, `variant_type` | Standard/robust designation and transformation type. |
+| `domain`, `source_dataset`, `source_sample_id`, `base_query_id` | Task and source identifiers. |
+| `model_ids` | Candidate-model order for all outcome arrays. |
+| `performance` | Source-task utility in [0, 1]; unsuccessful source outcomes are zero. |
+| `cost_cny` | Cost in CNY per request. |
+| `latency_ms` | Recorded latency in milliseconds. |
+| `embedding_model`, `embedding` | Encoder identity and the shared 1,024-dimensional feature vector. |
+| `media_json` | Media archive, image paths, and image provenance. |
+| `provenance_json`, `source_split`, `is_validation` | Source metadata and split annotations. |
+`performance[i]`, `cost_cny[i]`, and `latency_ms[i]` correspond to `model_ids[i]`.
+USD cost proxies use the fixed benchmark conversion **1 USD = 7.2 CNY**.
+Candidate identities are listed in [metadata/](metadata/).
+### Request features
+General Text and Chinese use BGE-M3 features. Multimodal uses
+[CLIP ViT-B/32][radford2021clip] (`openai/clip-vit-base-patch32`, checkpoint
+`3d74acf9a28c67741b2f4f2ea7635f0aaf6f0268`). Full text is split into chunks
+of up to 75 payload tokens. Normalized text vectors are pooled by payload-token
+count; normalized vectors for all request images are pooled equally. Each
+modality pool is normalized, then the two 512-vectors are concatenated and divided
+by the square root of two. The result is a unit-norm 1,024-dimensional vector.
+All methods and fitting seeds share these features. Reference answers are not
+inputs to the encoder.
+Router fitting and prediction replay use the bundled features without downloading
+encoder weights or images. Original images and reference annotations are supplied
+in the [multimodal media archives](media/multimodal/README.md), organized as
+`train.zip`, `validation.zip`, `standard_test.zip`, and `robust_test.zip`.
+## Benchmark composition
+General Text covers code, general reasoning, knowledge QA, mathematical reasoning,
+reading understanding, and long context. Its robust requests apply semantic
+rewrites, language variation, and long-context noise to source requests.
+Multimodal training draws from 18 [MMR-Bench-V2][ma2026mmrbench] sub-benchmarks.
+Standard evaluation covers [VQAv2][goyal2017vqav2], [AI2D][kembhavi2016ai2d],
+[ScienceQA and ScienceQA-IMG][lu2022scienceqa], [DocVQA][mathew2021docvqa],
+and [OCR-VQA][mishra2019ocrvqa]. Robust evaluation includes text rewriting,
+context noise, image variation, and combined perturbations. The standard and
+robust populations are independently constructed; their comparison is at the
+distribution level.
+Chinese covers [AlignBench][liu2023alignbench], [C-Eval][huang2023ceval],
+[CIF-Bench][li2024cifbench], [CLUE tasks][xu2020clue], [CMB-Clin][wang2023cmb],
+[Chinese SimpleQA][he2025chinesesimpleqa], [FinEval][guo2023fineval],
+[FollowBench-zh][jiang2024followbench], [GAOKAO-Bench][zhang2023gaokao],
+[LawBench][fei2023lawbench], and [Chinese LongBench tasks][bai2023longbench].
+Its robust requests use semantic rewriting, English translation, and prefix noise.
+### Robust test populations
+| Scenario | Transformation | Test requests |
+| --- | --- | ---: |
+| General Text | Semantic rewrite | 963 |
+| General Text | Language variation | 964 |
+| General Text | Long-context noise | 967 |
+| Multimodal | Text rewrite | 475 |
+| Multimodal | Context noise | 475 |
+| Multimodal | Image variation | 475 |
+| Multimodal | Combined perturbation | 475 |
+| Chinese | Semantic rewrite | 667 |
+| Chinese | English translation | 666 |
+| Chinese | Prefix noise | 667 |
+Transformations preserve the answer contract and the source scorer. Image
+variation includes JPEG degradation, cutout, Gaussian noise, Gaussian blur, and
+contrast reduction. Text-scenario variants retain their source identifiers.
+## Compared methods
+The report compares **13 implementations** using [ORBIT][lai2026orbit] interfaces,
+with the same candidate pool, request features, and test requests within each
+scenario. Mio selector is fitted separately per scenario and configured using
+validation requests.
+| Implementation | Method reference |
 | --- | --- |
+| Mio selector | Technical report Section 4.2 |
 | Avengers | [The Avengers][zhang2025avengers] |
 | Avengers-Pro | [Performance-Efficiency Optimized Routing][zhang2025avengerspro] |
+| Causal RM-Softmax | [Causal LLM Routing][tsiourvas2025causalrouting] |
 | Eagle | [Eagle][zhao2024eagle] |
 | EmbedLLM-RouteJudge | [EmbedLLM][zhuang2024embedllm] |
 | GraphRouter | [GraphRouter][feng2024graphrouter] |
 | HybridLLM | [Hybrid LLM][ding2024hybridllm] |
+| kNN | [TensorOpera][stripelis2024tensoropera] / [ORBIT][lai2026orbit] |
+| MLP | [TensorOpera][stripelis2024tensoropera] / [ORBIT][lai2026orbit] |
+| MIRT | [IRT-Router][song2025irtrouter] |
+| RouteLLM-MF | [RouteLLM][ong2024routellm] |
+| SVM Router | [TensorOpera][stripelis2024tensoropera] / [ORBIT][lai2026orbit] |
+The comparison concerns the offline request-level selector with \\(T=1\\).
+Implementation settings are provided with the prediction files and in the code.
+## Metrics
+Evaluate 100 evenly spaced nominal budgets between the cheapest and most expensive
+fixed-model mean costs on each reporting population. Predicted cost determines
+feasibility; predicted utility selects among candidates. A request with no feasible
+candidate uses the model with the lowest predicted cost. Observed outcomes
+determine mean utility \\(Q_\pi(C)\\) and realized mean cost \\(\bar c_\pi(C)\\).
+For the common nominal-budget grid \\(\mathcal G\\), peak quality and threshold cost are:
 $$
 \begin{aligned}
 \end{aligned}
 $$
+Cost@90, Cost@95, and Cost@100 use \\(\tau=0.90,0.95,1.00\\), respectively.
+An unattained target is **N/R**. Savings relative to the strongest fixed model are:
 $$
 \operatorname{Saving}_{\pi}\mathbin{@}\tau=1-\frac{\bar{c}_{\pi}\mathbin{@}\tau}{\bar c_{\mathrm{strongest}}}.
 $$
+nAUC integrates the cumulative-maximum quality envelope over increasing nominal
+budget and divides by the full budget interval:
 $$
+\begin{aligned}
+Q_\pi^{\mathrm{env}}(C^{(g)})&=\max_{1\leq h\leq g}Q_\pi(C^{(h)}),\\
+\operatorname{nAUC}(\pi)&=\frac{1}{C^{(K)}-C^{(1)}}\sum_{g=1}^{K-1}
 \frac{Q_\pi^{\mathrm{env}}(C^{(g)})+Q_\pi^{\mathrm{env}}(C^{(g+1)})}{2}
 \bigl(C^{(g+1)}-C^{(g)}\bigr).
+\end{aligned}
 $$
+Curves contain evaluated selector points only. The quality-cost figure uses
+realized cost on a logarithmic axis; its geometric area is not nAUC.
+Macro nAUC equally averages the six scenario/standard-robust scores.
+Latency is reported as an auxiliary resource measure.
+Multimodal uses fitting seeds 20260917, 20260918, and 20260919. Curves are averaged
+pointwise on the common budget grid **before** computing their metrics. Reported
+uncertainties are exact run-bootstrap standard errors over all 27 ordered
+resamples. General Text and Chinese each use seed 20260917.
+Dataset costs use CNY. Report tables use USD for General Text and Multimodal,
+and CNY for Chinese; divide the first two scenarios' `cost_cny` values by 7.2
+for those reporting units. Multimodal costs are fixed proxies.
+## Results
+Mio selector achieves the highest macro nAUC, **0.6360**, followed by MLP at
+**0.6183**. The complete tables report every method and each robustness group.
+![Quality-cost results for Mio selector and the 12 baselines across all six test populations.](assets/quality-cost-results.png)
+*Green: Mio selector frontier and peak. Hollow circles: baseline peaks. Blue
+diamond: best baseline by nAUC. Multimodal points use the mean curve of three
+fits. Costs are USD for General Text and Multimodal, and CNY for Chinese.*
+Costs below are per request. **N/R** denotes an unattained aggregate-curve target;
+`(r/3)` records the number of individual Multimodal runs reaching that target.
 <!-- BEGIN GENERATED RESULTS -->
 ### Macro nAUC
 | Rank | Implementation | Macro nAUC |
 | Chinese | Semantic rewrite | 0.6156 | 5/13 | 0.6086 | 4/13 |
 | Chinese | English translation | 0.5893 | 7/13 | 0.5802 | 5/13 |
 | Chinese | Prefix noise | 0.6375 | 7/13 | 0.6287 | 5/13 |
 <!-- END GENERATED RESULTS -->
+## Reproduce the report
+From the [GitHub repository](https://github.com/SinapisAI/collective-model-intelligence):
+```sh
+python -m pip install -e '.[report]'
+python -m examples.reproduce_report --mode replay --output runs/report-replay
+python -m examples.reproduce_report --mode fit --device cpu --output runs/report-fit
+```
+Replay evaluates the **65 supplied prediction arrays across 13 methods** and
+checks the aggregate against [report/aggregate-results.json](report/aggregate-results.json).
+Fit trains the 12 baseline implementations; Mio is supplied through its prediction
+arrays. Predictions are aligned by request and model identifiers. All requests
+in the official test split contribute to evaluation.
+[report/reproduction-manifest.json](report/reproduction-manifest.json) records
+prediction checksums and fitting seeds. [manifest.json](manifest.json) records
+data and media checksums. The original benchmark questions and images retain
+their upstream attribution and terms in [LICENSES.md](LICENSES.md).
 [bai2023longbench]: https://arxiv.org/abs/2308.14508
 [chen2024routerdc]: https://proceedings.neurips.cc/paper_files/paper/2024/hash/7a641b8ec86162fc875fb9f6456a542f-Abstract-Conference.html
 [zhang2025modelsat]: https://ojs.aaai.org/index.php/AAAI/article/view/34790
 [zhao2024eagle]: https://arxiv.org/abs/2409.15518
 [zhuang2024embedllm]: https://proceedings.iclr.cc/paper_files/paper/2025/hash/bf5e4b85d203481d6e37bd32d9600162-Abstract-Conference.html
+[radford2021clip]: https://proceedings.mlr.press/v139/radford21a.html

data/chinese/test-00000-of-00001.parquet CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:427a198c55d8a323d4f525506cc380d0c3bf45ef6f483c56411be8119e521984
-size 103944022

 version https://git-lfs.github.com/spec/v1
+oid sha256:22761d0acfeb54e557dcd3db18fafe5d034e117ee1728bcd7696857e1942ab6f
+size 84862750

data/multimodal/test-00000-of-00001.parquet CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4751c0bb642aa37894732c974f06b57e4f91747a22fcbf9469954595a374b93b
-size 27595772

 version https://git-lfs.github.com/spec/v1
+oid sha256:02df700092270b1c6f3dde221a78b705bbe68dd9d3daaa088f0c053eee4a3a44
+size 25798240

data/multimodal/validation-00000-of-00001.parquet CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f1c1d9986217e0f4dec261cee3ba08f6772f6c1fcb62b7b64a359c2e2290d010
-size 2622889

 version https://git-lfs.github.com/spec/v1
+oid sha256:0901654a1cc862c14d78a96c0b3520bbfec63195b4691c58ea0c4d5a29e0ae47
+size 2622862

data/text/test-00000-of-00001.parquet CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6d6d78f6fa62098a613c12404e95f72fafa850f558b409d68623d72cc54c7c4e
-size 289269468

 version https://git-lfs.github.com/spec/v1
+oid sha256:3a09438c80bc4ec00b48a6a180bea6abd54eb638d0ade344041bc81a0add090d
+size 263117422

manifest.json CHANGED Viewed

@@ -1,13 +1,13 @@
 {
-  "schema_version": "robust-route-bench-v1.1.0",
-  "version": "1.1.2-20261015",
-  "generated_at": "2026-10-15",
   "repository": "SinapisAI/RobustRouteBench",
   "fully_public": true,
   "self_contained": true,
   "external_data_dependencies": false,
-  "validation_overlaps_test": true,
-  "validation_policy": "400 standard + 100 robust per scenario",
   "cost_policy": {
     "currency": "CNY",
     "usd_to_cny_rate": 7.2,
@@ -18,88 +18,94 @@
   "media_archives": {
     "train": {
       "path": "media/multimodal/train.zip",
-      "bytes": 533630327,
-      "sha256": "c744383941c15d147ac481e584f31ce1629e2066d7740da8a8413d9fb6985f66"
     },
     "standard": {
       "path": "media/multimodal/standard_test.zip",
-      "bytes": 359143226,
-      "sha256": "4dee661c0f4a1247a1f33572707a3e4bfeeb21c31beee10fb361f498b11afcef"
     },
     "robust": {
       "path": "media/multimodal/robust_test.zip",
-      "bytes": 454272640,
-      "sha256": "0b64e70ac80792ff13b17b24dc06d72fccee1444592e0a3f24ca48eab456b831"
     }
   },
   "scenarios": {
     "text": {
-      "unique_queries": 10188,
       "models": 63,
-      "published_rows": {
         "train": 3000,
         "validation": 500,
-        "test": 7188
       },
-      "source_rows": {
-        "train": 3000,
-        "standard": 4188,
-        "robust": 3000
       },
       "validation_rows_by_type": {
         "standard": 400,
         "robust": 100
-      }
     },
     "multimodal": {
-      "unique_queries": 11000,
       "models": 45,
-      "published_rows": {
         "train": 4000,
         "validation": 500,
-        "test": 7000
       },
-      "source_rows": {
-        "train": 4000,
-        "standard": 5000,
-        "robust": 2000
       },
       "validation_rows_by_type": {
         "standard": 400,
         "robust": 100
-      }
     },
     "chinese": {
-      "unique_queries": 14090,
       "models": 10,
-      "published_rows": {
         "train": 7590,
         "validation": 500,
-        "test": 6500
       },
-      "source_rows": {
-        "train": 7590,
-        "standard": 4400,
-        "robust": 2100
       },
       "validation_rows_by_type": {
         "standard": 400,
         "robust": 100
-      }
     }
   },
   "files": {
     "LICENSES.md": {
-      "bytes": 1951,
-      "sha256": "25a37064f44d101f433d64cf449100abed88dfd11bfe9e20fcb87573f2ad193b"
     },
     "README.md": {
-      "bytes": 41856,
-      "sha256": "c9c5a37f61899703f9fa73dc044b94ecc72a1e77dfd2ebb7eb92317ab265a0d9"
     },
     "data/chinese/test-00000-of-00001.parquet": {
-      "bytes": 103944022,
-      "sha256": "427a198c55d8a323d4f525506cc380d0c3bf45ef6f483c56411be8119e521984"
     },
     "data/chinese/train-00000-of-00001.parquet": {
       "bytes": 56291524,
@@ -110,20 +116,20 @@
       "sha256": "273dcc0ac995cfc5077c0347b75e027dce518611cc78afaa9c34009dde6fcfc1"
     },
     "data/multimodal/test-00000-of-00001.parquet": {
-      "bytes": 27595772,
-      "sha256": "4751c0bb642aa37894732c974f06b57e4f91747a22fcbf9469954595a374b93b"
     },
     "data/multimodal/train-00000-of-00001.parquet": {
       "bytes": 15494283,
       "sha256": "2d0446d8f990a053c6960cd5917256786b91fc95a66cc67e290c553afeb04888"
     },
     "data/multimodal/validation-00000-of-00001.parquet": {
-      "bytes": 2622889,
-      "sha256": "f1c1d9986217e0f4dec261cee3ba08f6772f6c1fcb62b7b64a359c2e2290d010"
     },
     "data/text/test-00000-of-00001.parquet": {
-      "bytes": 289269468,
-      "sha256": "6d6d78f6fa62098a613c12404e95f72fafa850f558b409d68623d72cc54c7c4e"
     },
     "data/text/train-00000-of-00001.parquet": {
       "bytes": 23676804,
@@ -146,8 +152,8 @@
       "sha256": "a667c93e09ef001481690a64e691e8637c0fc25bc9537e1de19e3a71cec9723e"
     },
     "normalized-schema.json": {
-      "bytes": 2870,
-      "sha256": "1964859bb6bcc59d9a7098d35b43279c0e0fdd921b7dbb0f208e3fa62d176421"
     },
     ".gitattributes": {
       "bytes": 2504,
@@ -166,570 +172,567 @@
       "sha256": "f3bdccb183c6ef22b8c95cbd3231edb56636d2cb9d12a2ed06360241ecb384f3"
     },
     "release-audit.json": {
-      "bytes": 2433,
-      "sha256": "f74df6fe7ec0a23b4b9acd0ca8b7b065343299a84a9c12c291829cff1e43f980"
     },
     "report/aggregate-results.json": {
-      "bytes": 5185902,
-      "sha256": "1f1386ba63dda78286885bb4e50af1da0d9441056c4121cdf56d4c0fc7670de6"
     },
     "report/features/clip-fusion-features.npz": {
       "bytes": 38483446,
       "sha256": "735b413de957ebd0cb555f17c6e06dd74c4abab42fb43a7281e6e97282215065"
     },
     "report/predictions/chinese/seed20260917/avengers.json": {
-      "bytes": 106539,
-      "sha256": "6e14c4020b4bc4b6d90a80a3b15d2658e0f33806ddcbd19f9ca121f4a71371c3"
     },
     "report/predictions/chinese/seed20260917/avengers.npz": {
-      "bytes": 142401,
-      "sha256": "fcd8b31ee2ee72646f2ce92615c187afafe5c211bbad687c2548c1f4b3299e7a"
     },
     "report/predictions/chinese/seed20260917/avengerspro.json": {
-      "bytes": 112681,
-      "sha256": "8def9618efe5513efe856d4fa8367144d6962ea589478c981fa8736eae67e553"
     },
     "report/predictions/chinese/seed20260917/avengerspro.npz": {
-      "bytes": 266780,
-      "sha256": "df33ceb06775631501c05cb94e88678d1dbefda69efa49ed425135b96587e841"
     },
     "report/predictions/chinese/seed20260917/causal_rm_softmax.json": {
-      "bytes": 102895,
-      "sha256": "388c35be4264ee51639ff1c72163539cdcb254fad2455ea09b9fe79f4ef6be15"
     },
     "report/predictions/chinese/seed20260917/causal_rm_softmax.npz": {
-      "bytes": 719287,
-      "sha256": "5f32a2c48622299fac83beba10fce768e2f7f230a47a1581456015f8e44f9a06"
     },
     "report/predictions/chinese/seed20260917/eagle.json": {
-      "bytes": 116314,
-      "sha256": "cbf49ffe6c734e01b35b79c3242660989b14740c2003bf0c0bdaf81e0bb52e9e"
     },
     "report/predictions/chinese/seed20260917/eagle.npz": {
-      "bytes": 776117,
-      "sha256": "8757ffcb10bc0aaca23215d7c57a1d0c52e83109ef479c1f636363213b6cb7d6"
     },
     "report/predictions/chinese/seed20260917/embedllm_routejudge.json": {
-      "bytes": 101443,
-      "sha256": "033f6e963f09f1b475f19e26dbaa6a96d0ffaeb9ad5138ed6ac4508b8baa836c"
     },
     "report/predictions/chinese/seed20260917/embedllm_routejudge.npz": {
-      "bytes": 321990,
-      "sha256": "e0f5bce13e75893354af644612cf584ba0f90c62302e870acf7fd54cf701166a"
     },
     "report/predictions/chinese/seed20260917/graphrouter.json": {
-      "bytes": 102750,
-      "sha256": "7ae4c32f9867d4af2e9b7d567b155219e03f3d083b40fa4ca8d40f828d5bd449"
     },
     "report/predictions/chinese/seed20260917/graphrouter.npz": {
-      "bytes": 762504,
-      "sha256": "1aa4dde516ef3dd11632cf378ff1f3f37b8e8631fa7c672d05b07a756efb10a1"
     },
     "report/predictions/chinese/seed20260917/hybridllm.json": {
-      "bytes": 124022,
-      "sha256": "6ba0984884d09659719e210688e0528347dd634788361fba76125e77581377a9"
     },
     "report/predictions/chinese/seed20260917/hybridllm.npz": {
-      "bytes": 767849,
-      "sha256": "b658f6bbd5f7cc37291c7065a8163b21796fd7f92c8637dfde02806a0b40458c"
     },
     "report/predictions/chinese/seed20260917/knn.json": {
-      "bytes": 122411,
-      "sha256": "88d34dab13dec758f04986e54333fd0f4ff74d434f167c09b2ca254863bd1022"
     },
     "report/predictions/chinese/seed20260917/knn.npz": {
-      "bytes": 776732,
-      "sha256": "3100b7dbe440a6c51280cd18598feb127c2aca332492573cfdd5c892fa32c514"
     },
     "report/predictions/chinese/seed20260917/mio.json": {
-      "bytes": 116331,
-      "sha256": "7157f811bc3f17b804c611dac349c7ff5d50fb1255c676d3f43e824781569a7e"
     },
     "report/predictions/chinese/seed20260917/mio.npz": {
-      "bytes": 786641,
-      "sha256": "d86e56ac54aba21f9863eb55ca763059a4b8a536628be28338575282938aeb3b"
     },
     "report/predictions/chinese/seed20260917/mirt.json": {
-      "bytes": 106816,
-      "sha256": "eac6a559d611e167c226aa2e68c6babd33dd14ed6abfe3e5592b6ffb80628b3f"
     },
     "report/predictions/chinese/seed20260917/mirt.npz": {
-      "bytes": 736936,
-      "sha256": "6359fa117a93e85c6f79080343da6e90e801e193fa8d03b43437c9c601100784"
     },
     "report/predictions/chinese/seed20260917/mlp.json": {
-      "bytes": 115384,
-      "sha256": "5722ec5f4560077c62e4c094341fcf24febae54f66fa186c22bdac9ffb3ea841"
     },
     "report/predictions/chinese/seed20260917/mlp.npz": {
-      "bytes": 781325,
-      "sha256": "8571bc321a4ce6516a87ea1ed1b0a37f823b5239147ececcdbc1f6fd27b89706"
     },
     "report/predictions/chinese/seed20260917/routellm_mf.json": {
-      "bytes": 121624,
-      "sha256": "abf5f4e04c61da75ce209869cbfb8ad6cf91c1822ba2ce119835eefe22d64aad"
     },
     "report/predictions/chinese/seed20260917/routellm_mf.npz": {
-      "bytes": 768060,
-      "sha256": "53214ad4dfa02c64110a0c1b0e022f778de2c748d2c5ba8b51c094ab08e7b80f"
     },
     "report/predictions/chinese/seed20260917/svm_router.json": {
-      "bytes": 114888,
-      "sha256": "8f7140679731df84249cd1c00b72b6fe4aa7d4893e5e228fe2785ba7fe88562b"
     },
     "report/predictions/chinese/seed20260917/svm_router.npz": {
-      "bytes": 745694,
-      "sha256": "3bce58280e389fde06d9e7de49d4bb8d21664d9f0d42aaa655c3dbd763d18b80"
     },
     "report/predictions/multimodal/seed20260917/avengers.json": {
-      "bytes": 125416,
-      "sha256": "30e65ee34350d494871002cab007025e191a8900439ca51dea57843dfd697786"
     },
     "report/predictions/multimodal/seed20260917/avengers.npz": {
-      "bytes": 154891,
-      "sha256": "d138b48dec24dac0584aeda9422f8b3582705c3f4f503594b259536fd2e62d1e"
     },
     "report/predictions/multimodal/seed20260917/avengerspro.json": {
-      "bytes": 134656,
-      "sha256": "00f131568421c8e9cc1670474564c37473aead97e878bdc60c5e72015e0048ba"
     },
     "report/predictions/multimodal/seed20260917/avengerspro.npz": {
-      "bytes": 635725,
-      "sha256": "294fb104d04ba694c0b65720484cd7ac26480479f9dac3044c6f59afc4b08952"
     },
     "report/predictions/multimodal/seed20260917/causal_rm_softmax.json": {
-      "bytes": 123884,
-      "sha256": "210950e62f0094cae2bda34fb35e75336de158f7ac00346f2a84fb5449c7519a"
     },
     "report/predictions/multimodal/seed20260917/causal_rm_softmax.npz": {
-      "bytes": 3275540,
-      "sha256": "58235a8e779574b75efa8b1ba659aaaa55e7f9cfb6c7b18c4c3cf9742f228063"
     },
     "report/predictions/multimodal/seed20260917/eagle.json": {
-      "bytes": 138972,
-      "sha256": "b575bb28ff65f3488ff7cd5b38ba67a6720141bb10e4d9e1dd0ae83363b00a32"
     },
     "report/predictions/multimodal/seed20260917/eagle.npz": {
-      "bytes": 2381404,
-      "sha256": "55ea20e9814765107b17b1ee812dc101077a9872b2c0e0c755a215545df42385"
     },
     "report/predictions/multimodal/seed20260917/embedllm_routejudge.json": {
-      "bytes": 125098,
-      "sha256": "e601e36dc3f95149f3e9fdf5179bb921c0bb78d15b582ba76b0752a92171a06f"
     },
     "report/predictions/multimodal/seed20260917/embedllm_routejudge.npz": {
-      "bytes": 1116970,
-      "sha256": "928f7a7d88af3eaa898b8f11e08151fbd370fe93a5408f7fe8f3384009125fc3"
     },
     "report/predictions/multimodal/seed20260917/graphrouter.json": {
-      "bytes": 124373,
-      "sha256": "ade91e606d3739d5ef5354a090e6bcd9fbd2e2edcd231988c45c36bd205a1507"
     },
     "report/predictions/multimodal/seed20260917/graphrouter.npz": {
-      "bytes": 3129288,
-      "sha256": "27501649c5b32ced9806ebf5a4a6a9e93a6b28b56fb215839cfd5eced8e598d9"
     },
     "report/predictions/multimodal/seed20260917/hybridllm.json": {
-      "bytes": 134930,
-      "sha256": "26726ccf1b6bf60e98954cb7036f0dbc31ceaa1ddbaa9fc2b6dc3b9a69458228"
     },
     "report/predictions/multimodal/seed20260917/hybridllm.npz": {
-      "bytes": 2042431,
-      "sha256": "f7f2be4834dc56b5e2522c1b585ea0533ab2b76814152a078bc7fd001b39f097"
     },
     "report/predictions/multimodal/seed20260917/knn.json": {
-      "bytes": 128078,
-      "sha256": "d818b1bfcd7a1ac678fa4ea222c3dd24a966e10fe21b6f874f769080ad00c132"
     },
     "report/predictions/multimodal/seed20260917/knn.npz": {
-      "bytes": 2413849,
-      "sha256": "fa825cf4abdefd016b186eb3ca5874bce87c7f58baa3d460fb9450d71f36cd7a"
     },
     "report/predictions/multimodal/seed20260917/mio.json": {
-      "bytes": 123640,
-      "sha256": "8ebc2a5503aa106f2a380d15166eb98b144903fe608d6e57cf0a692c16cf0e32"
     },
     "report/predictions/multimodal/seed20260917/mio.npz": {
-      "bytes": 1230505,
-      "sha256": "0495e163cc72f238f051688dced8cdd4fc513d18d6927b9df1251d98844c67d5"
     },
     "report/predictions/multimodal/seed20260917/mirt.json": {
-      "bytes": 118476,
-      "sha256": "067d9df8b757df50c5bcf292e2985221bc9433e14f9d639c84f7067e21ec6067"
     },
     "report/predictions/multimodal/seed20260917/mirt.npz": {
-      "bytes": 3069339,
-      "sha256": "4741bafd50954f0f9c6e079f88166041ab1e1b5a715438b4cf574418cd121ed9"
     },
     "report/predictions/multimodal/seed20260917/mlp.json": {
-      "bytes": 125167,
-      "sha256": "269bf4838dd65095add6a9723a56dac700c1f7ab8c024fa3fbe753229c7d231b"
     },
     "report/predictions/multimodal/seed20260917/mlp.npz": {
-      "bytes": 3277484,
-      "sha256": "960fd3e8d44ef13ca09a89ccf8c2215b9631a9c1fbdabcb0079520e16c998e1e"
     },
     "report/predictions/multimodal/seed20260917/routellm_mf.json": {
-      "bytes": 127930,
-      "sha256": "d3b1d2032dcd4bb043239adfda9454e9f808e4d3e00ee16a1692d02dee783d0f"
     },
     "report/predictions/multimodal/seed20260917/routellm_mf.npz": {
-      "bytes": 2185117,
-      "sha256": "76bbeef42e1eecba46a513baaf1c2340d44758b7a8160ff3d75aaa4977145df9"
     },
     "report/predictions/multimodal/seed20260917/svm_router.json": {
-      "bytes": 124348,
-      "sha256": "f1b88ef04ed60b5b658a0720ce95f772676b5f0294f3ae88db849288bfd1c5ed"
     },
     "report/predictions/multimodal/seed20260917/svm_router.npz": {
-      "bytes": 2510268,
-      "sha256": "ed3245fd11143ee8d12e15ca7084225bf1e9674029113c45e124be58aeeb5e51"
     },
     "report/predictions/multimodal/seed20260918/avengers.json": {
-      "bytes": 124999,
-      "sha256": "b42de148b93aa633cdd9f815ebb39007c9d91eabe92f731c68401c6cec0df748"
     },
     "report/predictions/multimodal/seed20260918/avengers.npz": {
-      "bytes": 156598,
-      "sha256": "e764a1cf3a67a9158f708e07b8cc8022500f4750f6421a597f636dd1ec01dd0b"
     },
     "report/predictions/multimodal/seed20260918/avengerspro.json": {
-      "bytes": 128340,
-      "sha256": "a58ed48368e6e9afb6c05a51dcf04b1c212f3fbc4d12c506765d1791f357ac35"
     },
     "report/predictions/multimodal/seed20260918/avengerspro.npz": {
-      "bytes": 626949,
-      "sha256": "ba6163472beab5c2d10eae105600b572abddae22ff5dd556d417c8750384e9f1"
     },
     "report/predictions/multimodal/seed20260918/causal_rm_softmax.json": {
-      "bytes": 123405,
-      "sha256": "1f3baa33f5dda82757e4ef2a03aa4389751f3d90bbdc8b4dccf8a25bf7cf169c"
     },
     "report/predictions/multimodal/seed20260918/causal_rm_softmax.npz": {
-      "bytes": 3269945,
-      "sha256": "8a893141b43a3277e0fc9d0d6a321255a7a2fe7390daa519b8148cf31427afd5"
     },
     "report/predictions/multimodal/seed20260918/eagle.json": {
-      "bytes": 138972,
-      "sha256": "519189cd873010e67c168a509a045f49d8935a298051136d9ea6f2cd6c04b8f0"
     },
     "report/predictions/multimodal/seed20260918/eagle.npz": {
-      "bytes": 2381404,
-      "sha256": "55ea20e9814765107b17b1ee812dc101077a9872b2c0e0c755a215545df42385"
     },
     "report/predictions/multimodal/seed20260918/embedllm_routejudge.json": {
-      "bytes": 125043,
-      "sha256": "9d955920fa9a5085f62664967a8b75953ac6dd94976253e0eadcea21fef5295e"
     },
     "report/predictions/multimodal/seed20260918/embedllm_routejudge.npz": {
-      "bytes": 1113585,
-      "sha256": "d3bf73f6acf6a76373ca644ad6340085fe9c56dea1421b36d2b223bf71b8ce2d"
     },
     "report/predictions/multimodal/seed20260918/graphrouter.json": {
-      "bytes": 124105,
-      "sha256": "a9def29ff35af477dba15cc38a7bb1956b3fea2e0476ddebb3b0a9d02b6099e5"
     },
     "report/predictions/multimodal/seed20260918/graphrouter.npz": {
-      "bytes": 3112037,
-      "sha256": "2f9f78432164b6b82bdd7a7d09e35a2f2a83c523eca4fe48c340e3ba4f8e9658"
     },
     "report/predictions/multimodal/seed20260918/hybridllm.json": {
-      "bytes": 134930,
-      "sha256": "e9c56166141ef49f0fd2c5d8155e40d5eb7e6f18db9024ee343b5d63080908f0"
     },
     "report/predictions/multimodal/seed20260918/hybridllm.npz": {
-      "bytes": 2042431,
-      "sha256": "f7f2be4834dc56b5e2522c1b585ea0533ab2b76814152a078bc7fd001b39f097"
     },
     "report/predictions/multimodal/seed20260918/knn.json": {
-      "bytes": 128078,
-      "sha256": "2d57442c0709a14193324fdaaed7f4f033e9ee5cf8a36eac3ada424fc3525be1"
     },
     "report/predictions/multimodal/seed20260918/knn.npz": {
-      "bytes": 2413849,
-      "sha256": "fa825cf4abdefd016b186eb3ca5874bce87c7f58baa3d460fb9450d71f36cd7a"
     },
     "report/predictions/multimodal/seed20260918/mio.json": {
-      "bytes": 123862,
-      "sha256": "1437ff159779b4456051702b3b32ad39fe068fa6026cb688e616dbe6c1ee7112"
     },
     "report/predictions/multimodal/seed20260918/mio.npz": {
-      "bytes": 1235674,
-      "sha256": "fd5f777ef128a9260c73a33ad277ebe22112b0cf124394d50853c7c505f96084"
     },
     "report/predictions/multimodal/seed20260918/mirt.json": {
-      "bytes": 118003,
-      "sha256": "9f28544a7a827180afaba1dc4c2f1d99185ae3df3c294333d3b2e8b2c262135f"
     },
     "report/predictions/multimodal/seed20260918/mirt.npz": {
-      "bytes": 3059702,
-      "sha256": "6947410d98ea3360670ede2217d537d19f872055fb601ccf7c9267ccf6f8bf85"
     },
     "report/predictions/multimodal/seed20260918/mlp.json": {
-      "bytes": 125662,
-      "sha256": "20c96fe7a4ccb7d41186a45d9bcaa0e8b195c4f1123b50b974023ba74e67b217"
     },
     "report/predictions/multimodal/seed20260918/mlp.npz": {
-      "bytes": 3301899,
-      "sha256": "0e899491b0e3b75cea8ddf0435f0f5e6613538fbd7eb00ee8cff8417c81ac1ed"
     },
     "report/predictions/multimodal/seed20260918/routellm_mf.json": {
-      "bytes": 127930,
-      "sha256": "2a696e58a1ddf9697ffef67a004f58e76b8231bed0f2411db45eef7492b9fc29"
     },
     "report/predictions/multimodal/seed20260918/routellm_mf.npz": {
-      "bytes": 2185117,
-      "sha256": "76bbeef42e1eecba46a513baaf1c2340d44758b7a8160ff3d75aaa4977145df9"
     },
     "report/predictions/multimodal/seed20260918/svm_router.json": {
-      "bytes": 124283,
-      "sha256": "d61b009aeca034d0b71988917762440d68ac099d333df12a46379fb12e9100de"
     },
     "report/predictions/multimodal/seed20260918/svm_router.npz": {
-      "bytes": 2494195,
-      "sha256": "760f232c71bc7de11988c3ace31f36b510c6ee2ace218bc6c5fc9bf9ba58f173"
     },
     "report/predictions/multimodal/seed20260919/avengers.json": {
-      "bytes": 125876,
-      "sha256": "ad32137aa8e535912352bd14f768c77b94c97791079db80f0836e3ebe374a264"
     },
     "report/predictions/multimodal/seed20260919/avengers.npz": {
-      "bytes": 158804,
-      "sha256": "f11c349ea4108c9e43ce05c90c8c7427de030b4e933817406bc55cb666bba464"
     },
     "report/predictions/multimodal/seed20260919/avengerspro.json": {
-      "bytes": 128436,
-      "sha256": "b9f34d03b367113ec5056201ab59467c7dabd7e248960cb077aff80ec3ea5110"
     },
     "report/predictions/multimodal/seed20260919/avengerspro.npz": {
-      "bytes": 702527,
-      "sha256": "b71b48cf56b3aa60182d6712378f604eddfd5462632ebadf5fcb4a5bf329f064"
     },
     "report/predictions/multimodal/seed20260919/causal_rm_softmax.json": {
-      "bytes": 124057,
-      "sha256": "3f95201129357116b674ed47ffdc0be73325a981412b8faeb77a076dea1ddce9"
     },
     "report/predictions/multimodal/seed20260919/causal_rm_softmax.npz": {
-      "bytes": 3166118,
-      "sha256": "47357e6861cf0a646da48f3a781cd7de608d721a9dcd31894c77a0d067431740"
     },
     "report/predictions/multimodal/seed20260919/eagle.json": {
-      "bytes": 138972,
-      "sha256": "11d14a5e76203adf12d1c3d2f3a0cd1aef7b3180e6facadd7cae1c5375555eec"
     },
     "report/predictions/multimodal/seed20260919/eagle.npz": {
-      "bytes": 2381404,
-      "sha256": "55ea20e9814765107b17b1ee812dc101077a9872b2c0e0c755a215545df42385"
     },
     "report/predictions/multimodal/seed20260919/embedllm_routejudge.json": {
-      "bytes": 125296,
-      "sha256": "2938cf1bc86c1b523151e587690a83c892be075b9e6537515efb329e13258e48"
     },
     "report/predictions/multimodal/seed20260919/embedllm_routejudge.npz": {
-      "bytes": 1114610,
-      "sha256": "3ca1423981b8be12803b8d7f46c63757349114ef4b7c0bc319dd71dd3cafc358"
     },
     "report/predictions/multimodal/seed20260919/graphrouter.json": {
-      "bytes": 124897,
-      "sha256": "4d0c3520c5cdec306572adf463a7958686ae3ef99b91971d9840113f4779d974"
     },
     "report/predictions/multimodal/seed20260919/graphrouter.npz": {
-      "bytes": 3111664,
-      "sha256": "f71841a94c05e4103c13cda4eed89e184b7b9832b402e557f4c31c29cbeb0a52"
     },
     "report/predictions/multimodal/seed20260919/hybridllm.json": {
-      "bytes": 134930,
-      "sha256": "9badae11875fc401b21052025f46d95d9c92563fc6a0080ec2fae6cecc7541ea"
     },
     "report/predictions/multimodal/seed20260919/hybridllm.npz": {
-      "bytes": 2042431,
-      "sha256": "f7f2be4834dc56b5e2522c1b585ea0533ab2b76814152a078bc7fd001b39f097"
     },
     "report/predictions/multimodal/seed20260919/knn.json": {
-      "bytes": 128078,
-      "sha256": "e729189232ef659a1bd64d355945962ce336170acefb596f2bb004526125d098"
     },
     "report/predictions/multimodal/seed20260919/knn.npz": {
-      "bytes": 2413849,
-      "sha256": "fa825cf4abdefd016b186eb3ca5874bce87c7f58baa3d460fb9450d71f36cd7a"
     },
     "report/predictions/multimodal/seed20260919/mio.json": {
-      "bytes": 123756,
-      "sha256": "6340698f494580d68d12fa8493826d778518a1051981cc628b151b5d7db02f6f"
     },
     "report/predictions/multimodal/seed20260919/mio.npz": {
-      "bytes": 1232151,
-      "sha256": "10938b71e521f42cfa71acefc16f0e0ec84461ae22b51a81e9f5f42b424e70d4"
     },
     "report/predictions/multimodal/seed20260919/mirt.json": {
-      "bytes": 118181,
-      "sha256": "bf9086227ab765579dea4293ac21f0c1b9fafb05e09e9eb6af4e4bd6afe720e2"
     },
     "report/predictions/multimodal/seed20260919/mirt.npz": {
-      "bytes": 3077194,
-      "sha256": "927b6e1adcfe3d8a64cd7de9881ecfb0ff720f00cf75d7bf67b427f96008bf84"
     },
     "report/predictions/multimodal/seed20260919/mlp.json": {
-      "bytes": 123832,
-      "sha256": "8ec11d10c225bf47d3cd1f97d3a7277cc5e028ce2aabba3ee61609cae61599ed"
     },
     "report/predictions/multimodal/seed20260919/mlp.npz": {
-      "bytes": 3312756,
-      "sha256": "0a1fed68e51a646a513958f6b5a4d61a6afd7fe6b255901d9dcca0a72bfe36c7"
     },
     "report/predictions/multimodal/seed20260919/routellm_mf.json": {
-      "bytes": 127930,
-      "sha256": "966567e0fe4055340594dd723b3b8e34e748e9bfaaf03d211e7a1410e0408a06"
     },
     "report/predictions/multimodal/seed20260919/routellm_mf.npz": {
-      "bytes": 2185117,
-      "sha256": "76bbeef42e1eecba46a513baaf1c2340d44758b7a8160ff3d75aaa4977145df9"
     },
     "report/predictions/multimodal/seed20260919/svm_router.json": {
-      "bytes": 124499,
-      "sha256": "9efa35ba5ccc59e089a2910be818be0cda81a9032781370c3cbe4e45a2d9e557"
     },
     "report/predictions/multimodal/seed20260919/svm_router.npz": {
-      "bytes": 2540570,
-      "sha256": "bdfb8fc94cef61fbc51dfe6e6a20c99ee3a8195881e646f1398300c942b4cf52"
     },
     "report/predictions/text/seed20260917/avengers.json": {
-      "bytes": 103230,
-      "sha256": "4715ed2da3c57836fa7bf7a7f8c74fe843a947941179ddd75328e6ff22ca7127"
     },
     "report/predictions/text/seed20260917/avengers.npz": {
-      "bytes": 345519,
-      "sha256": "33ccae8c548d15f815072b0cbdbd8292004457728fbe0886eb9c34bea66e05b2"
     },
     "report/predictions/text/seed20260917/avengerspro.json": {
-      "bytes": 105316,
-      "sha256": "e56278f033ae0bf0e9f1366498d25db4260845cb24980ebd7076ec091ffe664a"
     },
     "report/predictions/text/seed20260917/avengerspro.npz": {
-      "bytes": 1392701,
-      "sha256": "b6dd7fb76865fd62644a3e31ac6bbc1ff269672f8565cf6a28b9eb60eb1bbfe6"
     },
     "report/predictions/text/seed20260917/causal_rm_softmax.json": {
-      "bytes": 99004,
-      "sha256": "8e63b5894d09b29be6915ca3b457921ccc5445f7b5cd3b14c72be80b80cf20d4"
     },
     "report/predictions/text/seed20260917/causal_rm_softmax.npz": {
-      "bytes": 3776964,
-      "sha256": "79422258eda8b38877ae58d7f0c6e6cc02fb1fe1d83954f9f8201e78c2cb00ca"
     },
     "report/predictions/text/seed20260917/eagle.json": {
-      "bytes": 110583,
-      "sha256": "39d0bd32921592ccaa3e086012454d4d42bb4a937602008d5804139de21ba738"
     },
     "report/predictions/text/seed20260917/eagle.npz": {
-      "bytes": 4417299,
-      "sha256": "6641f769e2a56544282ae1130b738185bf79004b3093801448deb00e35199503"
     },
     "report/predictions/text/seed20260917/embedllm_routejudge.json": {
-      "bytes": 101334,
-      "sha256": "56fd123dee2a2f13742e918d82543aaffe8f31f2c17525fe808a34ac1723f6df"
     },
     "report/predictions/text/seed20260917/embedllm_routejudge.npz": {
-      "bytes": 1649794,
-      "sha256": "fcccc4de6c53855b96b0a56e4eb5dee35f5df7760a6d24c77e20edccb194ce5f"
     },
     "report/predictions/text/seed20260917/graphrouter.json": {
-      "bytes": 102435,
-      "sha256": "c9b51c86d118dae24a9c191f3034a1389cdf2b50896e68d2dab19930032d2371"
     },
     "report/predictions/text/seed20260917/graphrouter.npz": {
-      "bytes": 3600991,
-      "sha256": "40b5194d34da4db7eb6350f1cac0415e8928856e7abba093219e085701ff447a"
     },
     "report/predictions/text/seed20260917/hybridllm.json": {
-      "bytes": 117322,
-      "sha256": "e74a3aeccca6302f3a2dbb94b7e9b1221d4994f431e7b53ddfccb07871b42d09"
     },
     "report/predictions/text/seed20260917/hybridllm.npz": {
-      "bytes": 4078259,
-      "sha256": "cb5f481e4b760fe686119e8bc3cac73e57e1069e3c96ff7a9f540ba8d2a102dc"
     },
     "report/predictions/text/seed20260917/knn.json": {
-      "bytes": 113840,
-      "sha256": "eaaa8d844d4d83b7c91b3621a1b5504417435f3c764afd32e301c00d88e37dd7"
     },
     "report/predictions/text/seed20260917/knn.npz": {
-      "bytes": 4399240,
-      "sha256": "0cbee52d58f530c8597589d72a43013905e8314e98e8c7ff9ab64907da92412f"
     },
     "report/predictions/text/seed20260917/mio.json": {
-      "bytes": 108732,
-      "sha256": "ddb209bb4181f4d3a0576d19001e33679d22f92294d989c9aa1e73d15e3f01c0"
     },
     "report/predictions/text/seed20260917/mio.npz": {
-      "bytes": 4755859,
-      "sha256": "42f74906309c2cb4ca15c8236cd391e2f19a4830509854252fa33a5136fde791"
     },
     "report/predictions/text/seed20260917/mirt.json": {
-      "bytes": 100000,
-      "sha256": "58ab293dcb47479efffdb6fed701a1d1992c83f5960f4e0cac9857b4dce91f5c"
     },
     "report/predictions/text/seed20260917/mirt.npz": {
-      "bytes": 3596905,
-      "sha256": "9420fdff881c83cac40683b7d4d5025c0e984f47bcd61358d07088e75cc5e47b"
     },
     "report/predictions/text/seed20260917/mlp.json": {
-      "bytes": 105281,
-      "sha256": "69736731d084fae5172aaf4b57a7922fb73f940107dade76d0d4dd6be43f7bde"
     },
     "report/predictions/text/seed20260917/mlp.npz": {
-      "bytes": 4650965,
-      "sha256": "336d9b39dd75feb0f113531c42265878f43947cd8bc32a3827400d2c881f0736"
     },
     "report/predictions/text/seed20260917/routellm_mf.json": {
-      "bytes": 110654,
-      "sha256": "5854b2e0aca1e347fdeebcb6dca099438b17790f4af4bd1d57e5847ae9b7f29a"
     },
     "report/predictions/text/seed20260917/routellm_mf.npz": {
-      "bytes": 4347159,
-      "sha256": "9be60b5bf315b25ae1f690711321683bccf10b7bece48aa33c31521c1abf70ca"
     },
     "report/predictions/text/seed20260917/svm_router.json": {
-      "bytes": 113071,
-      "sha256": "97c6fd4ffb85299cb67389db4603cae363e4455ac53bb8227aee83949950372d"
     },
     "report/predictions/text/seed20260917/svm_router.npz": {
-      "bytes": 3758138,
-      "sha256": "3464f7d976c7203de4bb4d286c50e5112490e8ab73d948a4f90650bb6c247385"
     },
     "report/reproduction-manifest.json": {
-      "bytes": 17585,
-      "sha256": "dede3662d0e5a528bcbf5ee85d8202e0eab96f2b0e9f47615a872ab8ccecbf59"
     },
     "media/multimodal/train.zip": {
-      "bytes": 533630327,
-      "sha256": "c744383941c15d147ac481e584f31ce1629e2066d7740da8a8413d9fb6985f66"
     },
     "media/multimodal/standard_test.zip": {
-      "bytes": 359143226,
-      "sha256": "4dee661c0f4a1247a1f33572707a3e4bfeeb21c31beee10fb361f498b11afcef"
     },
     "media/multimodal/robust_test.zip": {
-      "bytes": 454272640,
-      "sha256": "0b64e70ac80792ff13b17b24dc06d72fccee1444592e0a3f24ca48eab456b831"
     },
     "media/multimodal/manifest.json": {
-      "bytes": 924,
-      "sha256": "1d5e2f40a910da917641698d69be4d9a2f3665962b651360861713ac21998f2e"
     },
     "media/multimodal/README.md": {
-      "bytes": 1388,
-      "sha256": "f7ef4fd266647de976ec61cbc0790acef79713dfe73efa2a30048a5ec92ccc33"
     }
   },
-  "parent_revision": "c24982f3a47a960009f7040e70b220ab1c8dc068",
-  "score_repair": {
-    "successful_text_cells_restored": 20167,
-    "development_cells_restored": 522,
-    "failed_requests_score": 0,
-    "other_unverified_legacy_scores": "unchanged",
-    "all_outcomes_independently_regraded": false
-  },
   "embedding_policy": {
     "text": "BAAI/bge-m3",
     "chinese": "BAAI/bge-m3",

 {
+  "schema_version": "robust-route-bench-v2",
+  "version": "2.0.0",
+  "generated_at": "2026-10-18",
   "repository": "SinapisAI/RobustRouteBench",
   "fully_public": true,
   "self_contained": true,
   "external_data_dependencies": false,
+  "validation_overlaps_test": false,
+  "validation_policy": "500 configuration-selection requests per scenario: 400 standard and 100 robust",
   "cost_policy": {
     "currency": "CNY",
     "usd_to_cny_rate": 7.2,
   "media_archives": {
     "train": {
       "path": "media/multimodal/train.zip",
+      "requests": 4000,
+      "bytes": 482548115,
+      "sha256": "e1cee79b3fafcd8fa009cd939fc9bc6857b4543f81d6a99f565542de7bb31bc2"
+    },
+    "validation": {
+      "path": "media/multimodal/validation.zip",
+      "requests": 500,
+      "bytes": 55687379,
+      "sha256": "f33b8545ee7fa36dfe02b31401d2f7945e6c4ae55384c8367969d08a77fcde27"
     },
     "standard": {
       "path": "media/multimodal/standard_test.zip",
+      "requests": 4600,
+      "bytes": 332768099,
+      "sha256": "bc17bf5b918a50197959d0545f46703d36461cf69bf0dbf5af60553015ab3be3"
     },
     "robust": {
       "path": "media/multimodal/robust_test.zip",
+      "requests": 1900,
+      "bytes": 425408181,
+      "sha256": "9b9d9c35f9af70231445f42aeda1170996c767930399c1bd016f4c09b2181e53"
     }
   },
   "scenarios": {
     "text": {
       "models": 63,
+      "split_rows": {
         "train": 3000,
         "validation": 500,
+        "test": 6582
       },
+      "test_rows_by_type": {
+        "robust": 2894,
+        "standard": 3688
       },
       "validation_rows_by_type": {
         "standard": 400,
         "robust": 100
+      },
+      "total_instances": 10082
     },
     "multimodal": {
       "models": 45,
+      "split_rows": {
         "train": 4000,
         "validation": 500,
+        "test": 6500
       },
+      "test_rows_by_type": {
+        "robust": 1900,
+        "standard": 4600
       },
       "validation_rows_by_type": {
         "standard": 400,
         "robust": 100
+      },
+      "total_instances": 11000
     },
     "chinese": {
       "models": 10,
+      "split_rows": {
         "train": 7590,
         "validation": 500,
+        "test": 5900
       },
+      "test_rows_by_type": {
+        "robust": 2000,
+        "standard": 3900
       },
       "validation_rows_by_type": {
         "standard": 400,
         "robust": 100
+      },
+      "total_instances": 13990
     }
   },
   "files": {
     "LICENSES.md": {
+      "bytes": 1662,
+      "sha256": "83e00c2a8e201e13333fb9f7b1f004b7bca8956b9a414d6546f177816c72f70e"
     },
     "README.md": {
+      "bytes": 22630,
+      "sha256": "dbfa58218764913e99e4c5f14e7edf0f9b82079dcd6b7380667ab4446d8b437b"
     },
     "data/chinese/test-00000-of-00001.parquet": {
+      "bytes": 84862750,
+      "sha256": "22761d0acfeb54e557dcd3db18fafe5d034e117ee1728bcd7696857e1942ab6f"
     },
     "data/chinese/train-00000-of-00001.parquet": {
       "bytes": 56291524,
       "sha256": "273dcc0ac995cfc5077c0347b75e027dce518611cc78afaa9c34009dde6fcfc1"
     },
     "data/multimodal/test-00000-of-00001.parquet": {
+      "bytes": 25798240,
+      "sha256": "02df700092270b1c6f3dde221a78b705bbe68dd9d3daaa088f0c053eee4a3a44"
     },
     "data/multimodal/train-00000-of-00001.parquet": {
       "bytes": 15494283,
       "sha256": "2d0446d8f990a053c6960cd5917256786b91fc95a66cc67e290c553afeb04888"
     },
     "data/multimodal/validation-00000-of-00001.parquet": {
+      "bytes": 2622862,
+      "sha256": "0901654a1cc862c14d78a96c0b3520bbfec63195b4691c58ea0c4d5a29e0ae47"
     },
     "data/text/test-00000-of-00001.parquet": {
+      "bytes": 263117422,
+      "sha256": "3a09438c80bc4ec00b48a6a180bea6abd54eb638d0ade344041bc81a0add090d"
     },
     "data/text/train-00000-of-00001.parquet": {
       "bytes": 23676804,
       "sha256": "a667c93e09ef001481690a64e691e8637c0fc25bc9537e1de19e3a71cec9723e"
     },
     "normalized-schema.json": {
+      "bytes": 2863,
+      "sha256": "a343f52078df53e00c4425efd09fafbe90c6a02852ad19c0734f18f888851fe0"
     },
     ".gitattributes": {
       "bytes": 2504,
       "sha256": "f3bdccb183c6ef22b8c95cbd3231edb56636d2cb9d12a2ed06360241ecb384f3"
     },
     "release-audit.json": {
+      "bytes": 2224,
+      "sha256": "819b0624d6e4e44c7ce3c9dc0ed46dbc83c936dd5e3e1406179df29b185c40fe"
     },
     "report/aggregate-results.json": {
+      "bytes": 5185530,
+      "sha256": "66fc9d19e9aafa3d61c8ab3f3e52900b1acf07d4b41ac8167bf14a3780204735"
     },
     "report/features/clip-fusion-features.npz": {
       "bytes": 38483446,
       "sha256": "735b413de957ebd0cb555f17c6e06dd74c4abab42fb43a7281e6e97282215065"
     },
     "report/predictions/chinese/seed20260917/avengers.json": {
+      "bytes": 106547,
+      "sha256": "007909f8fec095f0fe7101292868676d51a94865d5ec197ce1e86baaacf3eea3"
     },
     "report/predictions/chinese/seed20260917/avengers.npz": {
+      "bytes": 131564,
+      "sha256": "fac73ad49bb6eb4365d50458e006b650dbd9aef782148e683839767d29db613f"
     },
     "report/predictions/chinese/seed20260917/avengerspro.json": {
+      "bytes": 112689,
+      "sha256": "24dce679784f1e132269b1197a0e340f8a1ac7b7289c587e55990f2056290947"
     },
     "report/predictions/chinese/seed20260917/avengerspro.npz": {
+      "bytes": 248661,
+      "sha256": "c6d4c99f3936800022af5bdcc345f496cef3f27046411a539002e381172b9144"
     },
     "report/predictions/chinese/seed20260917/causal_rm_softmax.json": {
+      "bytes": 102903,
+      "sha256": "5ea53118394543f3201709b9ab38bd1e71c2f74097879c2b6ca9bfee21f44127"
     },
     "report/predictions/chinese/seed20260917/causal_rm_softmax.npz": {
+      "bytes": 655113,
+      "sha256": "db1da64aaeeb5272adc963158870c04c5029bf4d30bc762e9ffaa856faa78f78"
     },
     "report/predictions/chinese/seed20260917/eagle.json": {
+      "bytes": 116322,
+      "sha256": "3ccba5eb97c021d97a97df3b8ebd53b7559ffa56ffe7dc1d728c56cd855d7d36"
     },
     "report/predictions/chinese/seed20260917/eagle.npz": {
+      "bytes": 705901,
+      "sha256": "5fbe1650e1728e89444f2a408bf763e50dc00def59dc223fd4b5032c156544a5"
     },
     "report/predictions/chinese/seed20260917/embedllm_routejudge.json": {
+      "bytes": 101451,
+      "sha256": "37ca674063928bdcdc622724f39e3ad932b074e48c24efb9944ea72fb856ac08"
     },
     "report/predictions/chinese/seed20260917/embedllm_routejudge.npz": {
+      "bytes": 294685,
+      "sha256": "9798166e4d0d2949379a1bdecb96dce0c02e15aa08a116e0c68a9bfdb040dae2"
     },
     "report/predictions/chinese/seed20260917/graphrouter.json": {
+      "bytes": 102758,
+      "sha256": "e5d0818a411721fde9d80e193645a6b0adb14844d4b2d3f3c9c42bd9af6b17f3"
     },
     "report/predictions/chinese/seed20260917/graphrouter.npz": {
+      "bytes": 694463,
+      "sha256": "e06d7fdfb162e6341b0462f82570764c28d59b28b1373af15dce9031ff7a01be"
     },
     "report/predictions/chinese/seed20260917/hybridllm.json": {
+      "bytes": 124030,
+      "sha256": "d3d1af9d3548b90639d7e2ce9ec5302f72970dbf26eac08ab1462747103685ab"
     },
     "report/predictions/chinese/seed20260917/hybridllm.npz": {
+      "bytes": 698950,
+      "sha256": "78137fcbc20154551a8a42079ddfcb30fe6a189f9c3c4befe5ac04cd0905b870"
     },
     "report/predictions/chinese/seed20260917/knn.json": {
+      "bytes": 122419,
+      "sha256": "ae7c36902f2362c0c14f4d6ddfa816250c74337206d38058ffbd1d61b2836172"
     },
     "report/predictions/chinese/seed20260917/knn.npz": {
+      "bytes": 706605,
+      "sha256": "ab4909a7e095dd1858839794271db91ffea805e015bdcecf199ef354259f917f"
     },
     "report/predictions/chinese/seed20260917/mio.json": {
+      "bytes": 116339,
+      "sha256": "cf6dc5eba4e287331c59f96196af6f290c018f07fc99aced5c2e685e89a287f7"
     },
     "report/predictions/chinese/seed20260917/mio.npz": {
+      "bytes": 714312,
+      "sha256": "861d3fd5110b55169ffcf4e1d6d099ebb982207320e4253a487a8f1f193c51be"
     },
     "report/predictions/chinese/seed20260917/mirt.json": {
+      "bytes": 106824,
+      "sha256": "a3e645f6c6088640f33ef73c288d9f5e7fae4d99183f1e63b441f8f336a9124d"
     },
     "report/predictions/chinese/seed20260917/mirt.npz": {
+      "bytes": 670605,
+      "sha256": "949034a807451176f04874305d7afc2ddaf38f7df8d97b79256235b9aa0fa62b"
     },
     "report/predictions/chinese/seed20260917/mlp.json": {
+      "bytes": 115392,
+      "sha256": "1acd5e0b9697961a61cafde48b18f245a58a92de77c45624ccbee4f3bf695c4c"
     },
     "report/predictions/chinese/seed20260917/mlp.npz": {
+      "bytes": 710721,
+      "sha256": "44bd905bc2e01d958d9e70e660c31ab1b1aee91b5dc65d3fbd02d61dfb398a22"
     },
     "report/predictions/chinese/seed20260917/routellm_mf.json": {
+      "bytes": 121632,
+      "sha256": "dd73f6e2a7b2b2c81214237c0704c3108aa94cb839eb74300c3d3b089e63782c"
     },
     "report/predictions/chinese/seed20260917/routellm_mf.npz": {
+      "bytes": 699205,
+      "sha256": "82a7228bce9282e30f928648312768d2f5b20b0ba4c487992263902c5046e536"
     },
     "report/predictions/chinese/seed20260917/svm_router.json": {
+      "bytes": 114896,
+      "sha256": "929a4f7936178e5c695d6c2b12589e471edb5f530dd9081608f4078bd84d7702"
     },
     "report/predictions/chinese/seed20260917/svm_router.npz": {
+      "bytes": 674338,
+      "sha256": "a14af06dd1213fccd6f5c9cd68ec24392c119a4766140018510102c2d3ee298b"
     },
     "report/predictions/multimodal/seed20260917/avengers.json": {
+      "bytes": 125424,
+      "sha256": "164458a46c3d1a0de9a6ee0efd6dcb248b305bd196b56c3dc5cb2eb3cc71d45b"
     },
     "report/predictions/multimodal/seed20260917/avengers.npz": {
+      "bytes": 147504,
+      "sha256": "56bb600409edecf4eda8c405ffade277da0e1d15fb324fe5b4a54c302c3e78cd"
     },
     "report/predictions/multimodal/seed20260917/avengerspro.json": {
+      "bytes": 134664,
+      "sha256": "49b718cf19d5bd47cd208a880a008ad9630e855c62e5d57d999a48b3912bbc89"
     },
     "report/predictions/multimodal/seed20260917/avengerspro.npz": {
+      "bytes": 599860,
+      "sha256": "3e759ac5882bdd02216d012017c71ff6f159fa2c8ac76fdad5d7142946dbe4b0"
     },
     "report/predictions/multimodal/seed20260917/causal_rm_softmax.json": {
+      "bytes": 123892,
+      "sha256": "8c1bc8844c80a57aae6b9348496972785f2536a0a3121b93907c80a7e4824256"
     },
     "report/predictions/multimodal/seed20260917/causal_rm_softmax.npz": {
+      "bytes": 3043127,
+      "sha256": "c5efff74f2d571f386a5a30f88fe99e3eba9742066a1cb5579e779c51d92093a"
     },
     "report/predictions/multimodal/seed20260917/eagle.json": {
+      "bytes": 138980,
+      "sha256": "69ff573786bcc801084b91789adfa2fbd0d4b9c23e1c1694f58b9635a87e24cd"
     },
     "report/predictions/multimodal/seed20260917/eagle.npz": {
+      "bytes": 2212904,
+      "sha256": "5b722e011383e4e6f59d4b16cef602af7c37fecd893fc95c1d15d753475e9b70"
     },
     "report/predictions/multimodal/seed20260917/embedllm_routejudge.json": {
+      "bytes": 125106,
+      "sha256": "e07ca92f26a431a6953b67986e0689c05fcbb1c87d6f76cbc5f9f1d741d8454d"
     },
     "report/predictions/multimodal/seed20260917/embedllm_routejudge.npz": {
+      "bytes": 1044836,
+      "sha256": "55ba38330457440076f416374f0626bd0d57c1bf1da19c41477d81ae288b238f"
     },
     "report/predictions/multimodal/seed20260917/graphrouter.json": {
+      "bytes": 124381,
+      "sha256": "401eccbd79b8bd9d913890ceb5ea21ec1cf4fb3c2a4d46a6360f7381f4b26d2e"
     },
     "report/predictions/multimodal/seed20260917/graphrouter.npz": {
+      "bytes": 2907915,
+      "sha256": "dc796cf3080b0c1f467e8fb1a453f10550c50a573d3465049dc2d7b92025a0a7"
     },
     "report/predictions/multimodal/seed20260917/hybridllm.json": {
+      "bytes": 134938,
+      "sha256": "b290a6f339eeeee295ac44829577485a429555e3f4a9c4f4409f9ffafae71a45"
     },
     "report/predictions/multimodal/seed20260917/hybridllm.npz": {
+      "bytes": 1898583,
+      "sha256": "bfdfb93ce851e5935ce7521b6e34e9d9aeb1d5d70cc712294ab4d8e9113f455f"
     },
     "report/predictions/multimodal/seed20260917/knn.json": {
+      "bytes": 128086,
+      "sha256": "77ad1a512367cec287f5cf983d70c50cef1265e275e7b1adcae54b0849caf2b3"
     },
     "report/predictions/multimodal/seed20260917/knn.npz": {
+      "bytes": 2243524,
+      "sha256": "f028ff9bb2d27d0f39897da1a997bf3f0fa9c62f76d2e7e5b67fa965c4119aba"
     },
     "report/predictions/multimodal/seed20260917/mio.json": {
+      "bytes": 123648,
+      "sha256": "447c68f896b067e80fadaa113ce675f966f7240cad9c23ae5f4c2627140ae5fc"
     },
     "report/predictions/multimodal/seed20260917/mio.npz": {
+      "bytes": 1143647,
+      "sha256": "870bd8fd663cf6f817514736b0bd73ee7afabbabe34c445db97078c3002ab958"
     },
     "report/predictions/multimodal/seed20260917/mirt.json": {
+      "bytes": 118484,
+      "sha256": "15c449359a5c8178845982f6bf8063be81b112c3ab0016a26734c5987fb906d3"
     },
     "report/predictions/multimodal/seed20260917/mirt.npz": {
+      "bytes": 2852604,
+      "sha256": "1d3bb466a75a9884f9f77b9989866113b434fc3075c397e4680caa2912bc9500"
     },
     "report/predictions/multimodal/seed20260917/mlp.json": {
+      "bytes": 125175,
+      "sha256": "6ccfddd375ec0cbb07d553ebbe1c69310f64d637a8084c8539b66048629f1dc1"
     },
     "report/predictions/multimodal/seed20260917/mlp.npz": {
+      "bytes": 3045351,
+      "sha256": "5b5f2d248ea1876520a7022ed50c88da09a7ce140c8ca72c2052b33a6986f68b"
     },
     "report/predictions/multimodal/seed20260917/routellm_mf.json": {
+      "bytes": 127938,
+      "sha256": "9aaa41da3ba564999ae40ec5eb9f4963f0781fd9a3811130836a5b970d48f3b8"
     },
     "report/predictions/multimodal/seed20260917/routellm_mf.npz": {
+      "bytes": 2030839,
+      "sha256": "d2ecb163d808fe4481fa50ebb926937d13bab58eaf49492edf3309a8a5eef252"
     },
     "report/predictions/multimodal/seed20260917/svm_router.json": {
+      "bytes": 124356,
+      "sha256": "f2d4990f543433b6170d037bceefe1ed18cc8fd463b30eb704eecf2e1c8874d2"
     },
     "report/predictions/multimodal/seed20260917/svm_router.npz": {
+      "bytes": 2539784,
+      "sha256": "741569e6b9c01ee8ff16a789d8d5b01477c6274407d255814554a24e81d024c3"
     },
     "report/predictions/multimodal/seed20260918/avengers.json": {
+      "bytes": 125007,
+      "sha256": "e7ef8f162313a6f608b15569f33dddc3ed0c41a7fad068facfd158cca23cd8f0"
     },
     "report/predictions/multimodal/seed20260918/avengers.npz": {
+      "bytes": 147399,
+      "sha256": "3d765ede4e689e2b65e931cc9f70e3d67c7173724c27dd2b3728418261a890ba"
     },
     "report/predictions/multimodal/seed20260918/avengerspro.json": {
+      "bytes": 128348,
+      "sha256": "0d9c3a585f0d0c47cb25bf656798e771429c61a306ecebf6f5bf17f644a62d13"
     },
     "report/predictions/multimodal/seed20260918/avengerspro.npz": {
+      "bytes": 593300,
+      "sha256": "051135680faadef0f6d742c27d5c805dc555743dca6265c6188855445eac5c46"
     },
     "report/predictions/multimodal/seed20260918/causal_rm_softmax.json": {
+      "bytes": 123413,
+      "sha256": "12d58751d637c7777324245fcc1cac6d67e86b7b2f048c4e76144af36df59dcf"
     },
     "report/predictions/multimodal/seed20260918/causal_rm_softmax.npz": {
+      "bytes": 3038256,
+      "sha256": "5e9808dc1f2cff137e7d61480067e3be7b64b9ed5ac920fae957be29b28a0ea7"
     },
     "report/predictions/multimodal/seed20260918/eagle.json": {
+      "bytes": 138980,
+      "sha256": "229a3e40a3478e19ce2293837d16496eb3949e91d641efa29023e3493436bfcb"
     },
     "report/predictions/multimodal/seed20260918/eagle.npz": {
+      "bytes": 2212904,
+      "sha256": "5b722e011383e4e6f59d4b16cef602af7c37fecd893fc95c1d15d753475e9b70"
     },
     "report/predictions/multimodal/seed20260918/embedllm_routejudge.json": {
+      "bytes": 125051,
+      "sha256": "3348d318cdbcce301fbc058aefacdf7208db4b5189958c0355fb074f8d2345fb"
     },
     "report/predictions/multimodal/seed20260918/embedllm_routejudge.npz": {
+      "bytes": 1041610,
+      "sha256": "1b723a8cd05ea68f40e60480c2089a102f8afbab3364228963028327aafe3cab"
     },
     "report/predictions/multimodal/seed20260918/graphrouter.json": {
+      "bytes": 124113,
+      "sha256": "36d0190ca17723705817fa3324f2f43599c61269f1a11152cbb6d5a8602cd70d"
     },
     "report/predictions/multimodal/seed20260918/graphrouter.npz": {
+      "bytes": 2891796,
+      "sha256": "5d87d4b983842f20818c3e78711e1c372104cc897b9ce3311b050bbf942bcf03"
     },
     "report/predictions/multimodal/seed20260918/hybridllm.json": {
+      "bytes": 134938,
+      "sha256": "ea9c93ca1fd0d0ba840437c3b50c9a50c588b187217ee7977bb296d1c86d4789"
     },
     "report/predictions/multimodal/seed20260918/hybridllm.npz": {
+      "bytes": 1898583,
+      "sha256": "bfdfb93ce851e5935ce7521b6e34e9d9aeb1d5d70cc712294ab4d8e9113f455f"
     },
     "report/predictions/multimodal/seed20260918/knn.json": {
+      "bytes": 128086,
+      "sha256": "eda333d491804a6e7cf209a748975dc7e5b3fe310138a9640b005dc5972d9858"
     },
     "report/predictions/multimodal/seed20260918/knn.npz": {
+      "bytes": 2243524,
+      "sha256": "f028ff9bb2d27d0f39897da1a997bf3f0fa9c62f76d2e7e5b67fa965c4119aba"
     },
     "report/predictions/multimodal/seed20260918/mio.json": {
+      "bytes": 123870,
+      "sha256": "1f1034b09d2de4fe269939aaf8c0a1f846e00adb5453ec30479fe3f8e8a44787"
     },
     "report/predictions/multimodal/seed20260918/mio.npz": {
+      "bytes": 1148393,
+      "sha256": "8129221cf4b7039dd098b9ec4e7e093b322eb7e0e867446e48a9a5ddd6a43745"
     },
     "report/predictions/multimodal/seed20260918/mirt.json": {
+      "bytes": 118011,
+      "sha256": "12e358efece8b61d5e7cbdd0a9125c07eb31401ef6a481e9f83ecc788100571d"
     },
     "report/predictions/multimodal/seed20260918/mirt.npz": {
+      "bytes": 2843260,
+      "sha256": "e07a5ca6e359768f243e2fc20f24022d259c118b5d379d5917032663cdca3465"
     },
     "report/predictions/multimodal/seed20260918/mlp.json": {
+      "bytes": 125670,
+      "sha256": "37fa29314a4c6aab04959ba491cc34ecfa0921bb10d65a3c683e6373a48b5e48"
     },
     "report/predictions/multimodal/seed20260918/mlp.npz": {
+      "bytes": 3068091,
+      "sha256": "21dc30beaa0fd6adf14c11d05c7012d645eab8ba6b0eb47aa55ac5439432b1ed"
     },
     "report/predictions/multimodal/seed20260918/routellm_mf.json": {
+      "bytes": 127938,
+      "sha256": "9a1541eff4510a7c04171f8393985d8c4a4cff4768e7d35d4abac001fe35acfb"
     },
     "report/predictions/multimodal/seed20260918/routellm_mf.npz": {
+      "bytes": 2030839,
+      "sha256": "d2ecb163d808fe4481fa50ebb926937d13bab58eaf49492edf3309a8a5eef252"
     },
     "report/predictions/multimodal/seed20260918/svm_router.json": {
+      "bytes": 124291,
+      "sha256": "107d863e269892105323c70d6ba883539f5d2b9befe563790878d33abd93fac0"
     },
     "report/predictions/multimodal/seed20260918/svm_router.npz": {
+      "bytes": 2524861,
+      "sha256": "980a68980e8741c9c2ecdb322fc3c80dfa7612847dbb215a5719b5dc926b10f8"
     },
     "report/predictions/multimodal/seed20260919/avengers.json": {
+      "bytes": 125884,
+      "sha256": "0262225289744d971d39dd471f8f834588554d9232668ce0362ac17df2d570e0"
     },
     "report/predictions/multimodal/seed20260919/avengers.npz": {
+      "bytes": 149911,
+      "sha256": "cba3e9fe713bee7375452c5344e9da3210183492a143f0f6dfb5e8c71b290af8"
     },
     "report/predictions/multimodal/seed20260919/avengerspro.json": {
+      "bytes": 128444,
+      "sha256": "edaca7e8e8eb2be10b9ecf960ad35234cb5ae2290b156edbb12101a31280bf69"
     },
     "report/predictions/multimodal/seed20260919/avengerspro.npz": {
+      "bytes": 663168,
+      "sha256": "f0a5d65295937f1f9bbadaee07bb3b315af3a01e117bebdc6088bfc8dfd03f1d"
     },
     "report/predictions/multimodal/seed20260919/causal_rm_softmax.json": {
+      "bytes": 124065,
+      "sha256": "4678b7bfd78f505aa239cf74195d65dea7ecb841bc0b5831da3de1294a8cae85"
     },
     "report/predictions/multimodal/seed20260919/causal_rm_softmax.npz": {
+      "bytes": 2941383,
+      "sha256": "5a880cf4aa9749f866e6142b6da6f40982dafdb04b86414104f9f9b71299dcca"
     },
     "report/predictions/multimodal/seed20260919/eagle.json": {
+      "bytes": 138980,
+      "sha256": "0bee6469e38a2c925fada634be70b0c1941a9674395c3e38fc677e918dd1b4e1"
     },
     "report/predictions/multimodal/seed20260919/eagle.npz": {
+      "bytes": 2212904,
+      "sha256": "5b722e011383e4e6f59d4b16cef602af7c37fecd893fc95c1d15d753475e9b70"
     },
     "report/predictions/multimodal/seed20260919/embedllm_routejudge.json": {
+      "bytes": 125304,
+      "sha256": "a3ae3ba989d09657387221fc0d0eb42535305f5d6f4faad5f201198c43c1e389"
     },
     "report/predictions/multimodal/seed20260919/embedllm_routejudge.npz": {
+      "bytes": 1042544,
+      "sha256": "58c53b1053557a3c472f9b782a51cfa846fb099df0b59ca2a50e77fbe12466db"
     },
     "report/predictions/multimodal/seed20260919/graphrouter.json": {
+      "bytes": 124905,
+      "sha256": "873a0dc8e5815fa415a917baab811cb52956be081b12c86f7464325860927b06"
     },
     "report/predictions/multimodal/seed20260919/graphrouter.npz": {
+      "bytes": 2892002,
+      "sha256": "98389eea94b547b5e2333fecbdfb1349684a123c906638c03f93271902c145d8"
     },
     "report/predictions/multimodal/seed20260919/hybridllm.json": {
+      "bytes": 134938,
+      "sha256": "82b2541d1b232398839fba6630dacbd5c13c8c6863dfe404950caae0fc483738"
     },
     "report/predictions/multimodal/seed20260919/hybridllm.npz": {
+      "bytes": 1898583,
+      "sha256": "bfdfb93ce851e5935ce7521b6e34e9d9aeb1d5d70cc712294ab4d8e9113f455f"
     },
     "report/predictions/multimodal/seed20260919/knn.json": {
+      "bytes": 128086,
+      "sha256": "23f8188bf6035ba2e9391bd1f7da55f763b1180e19b86450d038bf6ae008a4d6"
     },
     "report/predictions/multimodal/seed20260919/knn.npz": {
+      "bytes": 2243524,
+      "sha256": "f028ff9bb2d27d0f39897da1a997bf3f0fa9c62f76d2e7e5b67fa965c4119aba"
     },
     "report/predictions/multimodal/seed20260919/mio.json": {
+      "bytes": 123764,
+      "sha256": "6d538665acf94a43a13343de72a5aed5b332877093f463a2dec0e8e435dc9ff1"
     },
     "report/predictions/multimodal/seed20260919/mio.npz": {
+      "bytes": 1145238,
+      "sha256": "44ae64729db23ff98929633272b88187606ce269eafebe5596e150c13e3f7eba"
     },
     "report/predictions/multimodal/seed20260919/mirt.json": {
+      "bytes": 118189,
+      "sha256": "045495a57627b44fff807874cdbf7f9a306c58b0e67a8a41b7d1be79f166473d"
     },
     "report/predictions/multimodal/seed20260919/mirt.npz": {
+      "bytes": 2859476,
+      "sha256": "2b288a46f3ca23e462c33dcee625cb67603c17a605c19cfa2dc30a4cca3eaab2"
     },
     "report/predictions/multimodal/seed20260919/mlp.json": {
+      "bytes": 123840,
+      "sha256": "71a171ad8849fe37679358fd7688d91fa986e97a56bfb4b66ca1dfde8fb62552"
     },
     "report/predictions/multimodal/seed20260919/mlp.npz": {
+      "bytes": 3077947,
+      "sha256": "9d0cbb5796665fe0d4b66a2cda2de254c6fa0a34a8cd3b841c3716f5adda82c8"
     },
     "report/predictions/multimodal/seed20260919/routellm_mf.json": {
+      "bytes": 127938,
+      "sha256": "544ee6eb9bfb02490f702a453a289a2420423880512bee0fa36b289a1df0035a"
     },
     "report/predictions/multimodal/seed20260919/routellm_mf.npz": {
+      "bytes": 2030839,
+      "sha256": "d2ecb163d808fe4481fa50ebb926937d13bab58eaf49492edf3309a8a5eef252"
     },
     "report/predictions/multimodal/seed20260919/svm_router.json": {
+      "bytes": 124507,
+      "sha256": "88e1061d55b965eba4ac06acf94b933b28bb07ee4bd18b2b83b638136232292f"
     },
     "report/predictions/multimodal/seed20260919/svm_router.npz": {
+      "bytes": 2569520,
+      "sha256": "8108f519ae9bdd81ef1bfd9bde72a9cf0ed81ab0cfa863b47288ff9398b63cf5"
     },
     "report/predictions/text/seed20260917/avengers.json": {
+      "bytes": 103238,
+      "sha256": "6e0a8eafd60141b9dd7f8713bcbf275d6fab1432aad4d640157640f4566a3e99"
     },
     "report/predictions/text/seed20260917/avengers.npz": {
+      "bytes": 324786,
+      "sha256": "d323b2ca2a72cfe511680447c1b2e64cf4ecf7637be5d57130777a58cf048dcb"
     },
     "report/predictions/text/seed20260917/avengerspro.json": {
+      "bytes": 105324,
+      "sha256": "e881d8c01866726f349ca7343154556ec6cda9d6f9bc0d963abb1a894f248e8e"
     },
     "report/predictions/text/seed20260917/avengerspro.npz": {
+      "bytes": 1276573,
+      "sha256": "65fd3fb6fad3f9ccd390cfa414a4c90157a5d88b2c9c8d22a769450de11eee10"
     },
     "report/predictions/text/seed20260917/causal_rm_softmax.json": {
+      "bytes": 99012,
+      "sha256": "8f95719b1b4d52dca913c76a49cd9317a7a4662c6c90f4cc85bb8ae9e7a5408b"
     },
     "report/predictions/text/seed20260917/causal_rm_softmax.npz": {
+      "bytes": 3452481,
+      "sha256": "951e633198248a35738832bb27efedf2b9a7a5e42d1281798b7bd5cd7147c2f5"
     },
     "report/predictions/text/seed20260917/eagle.json": {
+      "bytes": 110591,
+      "sha256": "be25fcb0f4a977503b09d6c57ba8b7343c5eff682089ea6a428cb620ab60dad9"
     },
     "report/predictions/text/seed20260917/eagle.npz": {
+      "bytes": 4047107,
+      "sha256": "85f0880c6de3c09fba3ba3fa996c91bb38c1fd1d5928ed55ee77814eff3bd598"
     },
     "report/predictions/text/seed20260917/embedllm_routejudge.json": {
+      "bytes": 101342,
+      "sha256": "875ed403238dbdcfff718c00cbc150b2a7f0fdd9805829e0f38f2fc3d1594569"
     },
     "report/predictions/text/seed20260917/embedllm_routejudge.npz": {
+      "bytes": 1520004,
+      "sha256": "2397353f2de8f1dbb94faabf1debde09aa9cc5520d3ad670fd0042033215b9bc"
     },
     "report/predictions/text/seed20260917/graphrouter.json": {
+      "bytes": 102443,
+      "sha256": "c79b1f1edcdd9dd1c90d2f27c6d37e2d0145e01871e519b916e4d1e2106b543a"
     },
     "report/predictions/text/seed20260917/graphrouter.npz": {
+      "bytes": 3298575,
+      "sha256": "7b3b999f51e1202e146eb55b5137fb5b2942c31f881fe82dd5d46c2c1bd812cb"
     },
     "report/predictions/text/seed20260917/hybridllm.json": {
+      "bytes": 117330,
+      "sha256": "f9a6f70c904adb0c9dc78717928c77b57a17264d6883bd00a9e894660576f205"
     },
     "report/predictions/text/seed20260917/hybridllm.npz": {
+      "bytes": 3732600,
+      "sha256": "7132f8e63711014f8dd3d5d49b3591baf865efc31d61a8961896c60b76b4cd21"
     },
     "report/predictions/text/seed20260917/knn.json": {
+      "bytes": 113848,
+      "sha256": "1e3705d97c6c203167333ed19bb2f49b9de783ee41eac343d8a9aa4068060352"
     },
     "report/predictions/text/seed20260917/knn.npz": {
+      "bytes": 4030855,
+      "sha256": "f5971af48d1f9fd8e0408c6025ecf74787b4163198648f5783a1d337b47a2899"
     },
     "report/predictions/text/seed20260917/mio.json": {
+      "bytes": 108740,
+      "sha256": "e6a1be078f7813a1861eda35fd7e2a02d4262212c6ecebd1f2ec6bd9a571924d"
     },
     "report/predictions/text/seed20260917/mio.npz": {
+      "bytes": 4354869,
+      "sha256": "1df6f5e58aa65cc7feb2e3589a88b451a6a80cd7de7f0712d34a78945615768e"
     },
     "report/predictions/text/seed20260917/mirt.json": {
+      "bytes": 100008,
+      "sha256": "b89bea9c239e08768de1fddffe3ee425d2f7e61e9ef96966bd48cdd526790d03"
     },
     "report/predictions/text/seed20260917/mirt.npz": {
+      "bytes": 3293973,
+      "sha256": "a84a271053012fd92840d97a3d80c6ea9194f83ea4caf605a1265660f79c8468"
     },
     "report/predictions/text/seed20260917/mlp.json": {
+      "bytes": 105289,
+      "sha256": "4baaeb6f6410b4c3a263fc869d0d479f2213426be2226fdebc50c2c60e3bbee1"
     },
     "report/predictions/text/seed20260917/mlp.npz": {
+      "bytes": 4257964,
+      "sha256": "8bf849d38a36f7e93b57d25edf0ba9039dbcf3c5c3bebd0a13271b4130993642"
     },
     "report/predictions/text/seed20260917/routellm_mf.json": {
+      "bytes": 110662,
+      "sha256": "915434d931cc4270edfaade7dcdcc5957cbd2f444db86cc0faee8af09f6c3c31"
     },
     "report/predictions/text/seed20260917/routellm_mf.npz": {
+      "bytes": 3979984,
+      "sha256": "beb729ee903c61437f9e4e0efa7f12271ad51fc111987087f5d0a4c9c5ea9300"
     },
     "report/predictions/text/seed20260917/svm_router.json": {
+      "bytes": 113079,
+      "sha256": "7106b4681e9d62de18b6743b6a3a65c29b9cb62e9097671b2c7d45fcb768474e"
     },
     "report/predictions/text/seed20260917/svm_router.npz": {
+      "bytes": 3705462,
+      "sha256": "b85f48e1fa13f3eb995eecfbbe6143e6eef73516b7cc0b5f3d9adfcaaaeca6a3"
     },
     "report/reproduction-manifest.json": {
+      "bytes": 17620,
+      "sha256": "a91b57270101fc27bcbf472eeea48b6aa61cc0158cf7d679da5043a04e9cac34"
     },
     "media/multimodal/train.zip": {
+      "bytes": 482548115,
+      "sha256": "e1cee79b3fafcd8fa009cd939fc9bc6857b4543f81d6a99f565542de7bb31bc2"
     },
     "media/multimodal/standard_test.zip": {
+      "bytes": 332768099,
+      "sha256": "bc17bf5b918a50197959d0545f46703d36461cf69bf0dbf5af60553015ab3be3"
     },
     "media/multimodal/robust_test.zip": {
+      "bytes": 425408181,
+      "sha256": "9b9d9c35f9af70231445f42aeda1170996c767930399c1bd016f4c09b2181e53"
     },
     "media/multimodal/manifest.json": {
+      "bytes": 891,
+      "sha256": "3c2f2bf2c06f12c4857476dbd20cc0d5a9ac65e0bb041e80224dcda3ee680662"
     },
     "media/multimodal/README.md": {
+      "bytes": 1180,
+      "sha256": "0b556d0e755aec9220bfc52b6f5bcc554338c5be5a463a07fbb2ef7617e2b6a7"
+    },
+    "media/multimodal/validation.zip": {
+      "bytes": 55687379,
+      "sha256": "f33b8545ee7fa36dfe02b31401d2f7945e6c4ae55384c8367969d08a77fcde27"
     }
   },
+  "parent_revision": "d05d92c57e13d2c38da4c9e822d8be456c56f24d",
   "embedding_policy": {
     "text": "BAAI/bge-m3",
     "chinese": "BAAI/bge-m3",

media/multimodal/README.md CHANGED Viewed

@@ -1,27 +1,24 @@
 # Multimodal media
-These archives contain the original images and accompanying question/source
-metadata for the Multimodal scenario of RobustRouteBench.
-| Archive | Role | Published requests |
 | --- | --- | ---: |
 | [train.zip](train.zip) | Training / 训练集 | 4,000 |
-| [standard_test.zip](standard_test.zip) | Standard test / 标准测试集 | 5,000 |
-| [robust_test.zip](robust_test.zip) | Robustness test / 鲁棒测试集 | 2,000 |
-Media downloads are **optional** when fitting or replaying routers using the
-published CLIP features. Download these archives to inspect images, recompute
-features, or evaluate visual models on the original requests.
-The counts above include the published validation overlay (400 standard and
-100 robust requests). Apply the source-group exclusions in the
-[dataset card](../../README.md#evaluation-populations-and-development-exclusion)
-for the report's final test populations (4,600 standard and 1,900 robust).
-Validation is an overlay, not a separate media archive.
-Each data row's `media_json.archive` points to one of these archives;
-`archive_member_path` or `archive_member_paths` identifies its image members.
-Keep archive-internal directories unchanged when resolving these references.
-[manifest.json](manifest.json) records archive sizes and SHA-256 checksums.
-See [LICENSES.md](../../LICENSES.md) for upstream attribution and terms.

 # Multimodal media
+| Archive | Role | Requests |
 | --- | --- | ---: |
 | [train.zip](train.zip) | Training / 训练集 | 4,000 |
+| [validation.zip](validation.zip) | Validation / 验证集 | 500 |
+| [standard_test.zip](standard_test.zip) | Standard test / 标准测试集 | 4,600 |
+| [robust_test.zip](robust_test.zip) | Robustness test / 鲁棒测试集 | 1,900 |
+Validation contains 400 standard and 100 robust requests for configuration
+selection. The test archives correspond directly to the technical report's
+Multimodal evaluation population.
+Images are optional when using the bundled CLIP features to fit or evaluate
+routers. Download the archives to inspect images, compute features, or run
+visual models on the benchmark requests.
+Each archive provides `instances.jsonl` with prompts and reference annotations,
+the required images, and a checksum manifest. Its `query_id` matches the Parquet
+data. A row's `media_json.archive` names the archive; `archive_member_path` or
+`archive_member_paths` identifies the corresponding image files.
+See [manifest.json](manifest.json) for archive checksums and
+[LICENSES.md](../../LICENSES.md) for upstream attribution and terms.

media/multimodal/manifest.json CHANGED Viewed

@@ -1,24 +1,29 @@
 {
-  "self_contained": true,
-  "external_download_required": false,
   "archives": {
     "train": {
       "path": "media/multimodal/train.zip",
-      "bytes": 533630327,
-      "sha256": "c744383941c15d147ac481e584f31ce1629e2066d7740da8a8413d9fb6985f66"
     },
     "standard": {
       "path": "media/multimodal/standard_test.zip",
-      "bytes": 359143226,
-      "sha256": "4dee661c0f4a1247a1f33572707a3e4bfeeb21c31beee10fb361f498b11afcef"
     },
     "robust": {
       "path": "media/multimodal/robust_test.zip",
-      "bytes": 454272640,
-      "sha256": "0b64e70ac80792ff13b17b24dc06d72fccee1444592e0a3f24ca48eab456b831"
     }
-  },
-  "train_archive_contents": "MMR-4000 instances, selection metadata, provenance, validation metadata, and all 4,285 referenced image files",
-  "standard_archive_contents": "Vision QA 5,000 instances and media",
-  "robust_archive_contents": "Robustness 2,000 instances, media, and references"
 }

 {
+  "repository": "SinapisAI/RobustRouteBench",
   "archives": {
     "train": {
       "path": "media/multimodal/train.zip",
+      "requests": 4000,
+      "bytes": 482548115,
+      "sha256": "e1cee79b3fafcd8fa009cd939fc9bc6857b4543f81d6a99f565542de7bb31bc2"
+    },
+    "validation": {
+      "path": "media/multimodal/validation.zip",
+      "requests": 500,
+      "bytes": 55687379,
+      "sha256": "f33b8545ee7fa36dfe02b31401d2f7945e6c4ae55384c8367969d08a77fcde27"
     },
     "standard": {
       "path": "media/multimodal/standard_test.zip",
+      "requests": 4600,
+      "bytes": 332768099,
+      "sha256": "bc17bf5b918a50197959d0545f46703d36461cf69bf0dbf5af60553015ab3be3"
     },
     "robust": {
       "path": "media/multimodal/robust_test.zip",
+      "requests": 1900,
+      "bytes": 425408181,
+      "sha256": "9b9d9c35f9af70231445f42aeda1170996c767930399c1bd016f4c09b2181e53"
     }
+  }
 }

media/multimodal/robust_test.zip CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:0b64e70ac80792ff13b17b24dc06d72fccee1444592e0a3f24ca48eab456b831
-size 454272640

 version https://git-lfs.github.com/spec/v1
+oid sha256:9b9d9c35f9af70231445f42aeda1170996c767930399c1bd016f4c09b2181e53
+size 425408181

media/multimodal/standard_test.zip CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4dee661c0f4a1247a1f33572707a3e4bfeeb21c31beee10fb361f498b11afcef
-size 359143226

 version https://git-lfs.github.com/spec/v1
+oid sha256:bc17bf5b918a50197959d0545f46703d36461cf69bf0dbf5af60553015ab3be3
+size 332768099

media/multimodal/train.zip CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c744383941c15d147ac481e584f31ce1629e2066d7740da8a8413d9fb6985f66
-size 533630327

 version https://git-lfs.github.com/spec/v1
+oid sha256:e1cee79b3fafcd8fa009cd939fc9bc6857b4543f81d6a99f565542de7bb31bc2
+size 482548115

media/multimodal/validation.zip ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:f33b8545ee7fa36dfe02b31401d2f7945e6c4ae55384c8367969d08a77fcde27
+size 55687379

normalized-schema.json CHANGED Viewed

@@ -1,5 +1,5 @@
 {
-  "schema_version": "robust-route-bench-v1.1.0",
   "layout": {
     "configs": [
       "text",
@@ -11,11 +11,12 @@
       "validation",
       "test"
     ],
-    "validation_overlaps_test": true,
     "test_slices": [
       "standard",
       "robust"
-    ]
   },
   "row_format": "one query with candidate-model-aligned arrays",
   "fields": {
@@ -45,8 +46,7 @@
   "cleaning_policy": {
     "non_successful_result": "performance is set to zero",
     "source_diagnostics_published": false,
-    "raw_model_responses_published": false,
-    "unverified_legacy_scores": "unchanged; not independently regraded"
   },
   "cost_policy": {
     "currency": "CNY",

 {
+  "schema_version": "robust-route-bench-v2",
   "layout": {
     "configs": [
       "text",
       "validation",
       "test"
     ],
+    "validation_overlaps_test": false,
     "test_slices": [
       "standard",
       "robust"
+    ],
+    "test_population": "paper evaluation requests; scored directly"
   },
   "row_format": "one query with candidate-model-aligned arrays",
   "fields": {
   "cleaning_policy": {
     "non_successful_result": "performance is set to zero",
     "source_diagnostics_published": false,
+    "raw_model_responses_published": false
   },
   "cost_policy": {
     "currency": "CNY",

release-audit.json CHANGED Viewed

@@ -1,71 +1,79 @@
 {
   "release_integrity_verified": true,
-  "validation_scope": "structure, finite ranges, overlays, source groups, frozen features",
   "all_outcomes_independently_regraded": false,
-  "unrecoverable_legacy_scores": "unchanged; no new ground-truth labels inferred",
   "scenarios": {
     "text": {
       "models": 63,
-      "training_rows": 3000,
-      "development_rows": 500,
-      "published_test_rows": 7188,
-      "final_standard_rows": 3688,
-      "final_robust_rows": 2894,
-      "training_evaluation_source_group_overlap": 0,
-      "development_final_test_source_group_overlap": 0,
-      "validation_overlay_identical_except_published_split": true,
       "scores_finite_and_unit_interval": true,
       "cost_latency_finite_nonnegative": true,
       "feature_width": 1024
     },
     "multimodal": {
       "models": 45,
-      "training_rows": 4000,
-      "development_rows": 500,
-      "published_test_rows": 7000,
-      "final_standard_rows": 4600,
-      "final_robust_rows": 1900,
-      "training_evaluation_source_group_overlap": 0,
-      "development_final_test_source_group_overlap": 0,
-      "validation_overlay_identical_except_published_split": true,
       "scores_finite_and_unit_interval": true,
       "cost_latency_finite_nonnegative": true,
       "feature_width": 1024
     },
     "chinese": {
       "models": 10,
-      "training_rows": 7590,
-      "development_rows": 500,
-      "published_test_rows": 6500,
-      "final_standard_rows": 3900,
-      "final_robust_rows": 2000,
-      "training_evaluation_source_group_overlap": 0,
-      "development_final_test_source_group_overlap": 0,
-      "validation_overlay_identical_except_published_split": true,
       "scores_finite_and_unit_interval": true,
       "cost_latency_finite_nonnegative": true,
       "feature_width": 1024
     }
-  },
-  "score_repair": {
-    "successful_text_cells_restored": 20167,
-    "development_cells_restored": 522,
-    "failed_requests_score": 0,
-    "other_unverified_legacy_scores": "unchanged",
-    "all_outcomes_independently_regraded": false
-  },
-  "known_failure_source_alignment": {
-    "text": {
-      "model_query_cells": 641844,
-      "known_failure_cells": 31018,
-      "nonzero_known_failure_cells": 0,
-      "private_public_score_mismatches": 0
-    },
-    "chinese": {
-      "model_query_cells": 140900,
-      "known_failure_cells": 79,
-      "nonzero_known_failure_cells": 0,
-      "private_public_score_mismatches": 0
-    }
   }
 }

 {
   "release_integrity_verified": true,
+  "validation_scope": "schema, numeric ranges, query identity, source-group partitions, features, and prediction alignment",
   "all_outcomes_independently_regraded": false,
   "scenarios": {
     "text": {
       "models": 63,
+      "split_rows": {
+        "train": 3000,
+        "validation": 500,
+        "test": 6582
+      },
+      "test_rows_by_type": {
+        "robust": 2894,
+        "standard": 3688
+      },
+      "validation_rows_by_type": {
+        "standard": 400,
+        "robust": 100
+      },
+      "total_instances": 10082,
+      "validation_test_query_overlap": 0,
+      "validation_test_source_group_overlap": 0,
+      "train_test_source_group_overlap": 0,
+      "train_validation_source_group_overlap": 0,
       "scores_finite_and_unit_interval": true,
       "cost_latency_finite_nonnegative": true,
       "feature_width": 1024
     },
     "multimodal": {
       "models": 45,
+      "split_rows": {
+        "train": 4000,
+        "validation": 500,
+        "test": 6500
+      },
+      "test_rows_by_type": {
+        "robust": 1900,
+        "standard": 4600
+      },
+      "validation_rows_by_type": {
+        "standard": 400,
+        "robust": 100
+      },
+      "total_instances": 11000,
+      "validation_test_query_overlap": 0,
+      "validation_test_source_group_overlap": 0,
+      "train_test_source_group_overlap": 0,
+      "train_validation_source_group_overlap": 0,
       "scores_finite_and_unit_interval": true,
       "cost_latency_finite_nonnegative": true,
       "feature_width": 1024
     },
     "chinese": {
       "models": 10,
+      "split_rows": {
+        "train": 7590,
+        "validation": 500,
+        "test": 5900
+      },
+      "test_rows_by_type": {
+        "robust": 2000,
+        "standard": 3900
+      },
+      "validation_rows_by_type": {
+        "standard": 400,
+        "robust": 100
+      },
+      "total_instances": 13990,
+      "validation_test_query_overlap": 0,
+      "validation_test_source_group_overlap": 0,
+      "train_test_source_group_overlap": 0,
+      "train_validation_source_group_overlap": 0,
       "scores_finite_and_unit_interval": true,
       "cost_latency_finite_nonnegative": true,
       "feature_width": 1024
     }
   }
 }

report/aggregate-results.json CHANGED Viewed

@@ -1,15 +1,16 @@
 {
   "schema_version": "routestressbench-report-v3",
-  "generated_at": "2026-10-15T07:02:20.742684+00:00",
   "protocol": {
-    "data_revision": "v1.1.0-20261015",
     "ranked_implementations": 13,
     "baseline_implementations": 12,
     "budget_points": 100,
     "auc_axis": "common nominal per-request cost budget",
     "auc_envelope": "cumulative maximum router quality over increasing budgets",
     "fixed_model_references_in_router_curve": false,
-    "development_group_overlap": 0,
     "multimodal_seeds": [
       20260917,
       20260918,
@@ -24,15 +25,7 @@
     "target_cost_estimator": "minimum realized mean cost on the pointwise mean curve",
     "target_reach_count": "number of individual runs reaching the target",
     "macro_std_identity": "descriptive sample standard deviation across six split scores",
-    "alias_audit_scope": "historical adapter audit; only 12 canonical baselines are ranked",
-    "score_repair": {
-      "successful_text_cells_restored": 20167,
-      "development_cells_restored": 522,
-      "failed_requests_score": 0,
-      "other_unverified_legacy_scores": "unchanged",
-      "all_outcomes_independently_regraded": false
-    },
-    "multimodal_features": "CLIP ViT-B/32 text/image late fusion"
   },
   "macro": {
     "Mio selector (offline)": {
@@ -135,7 +128,7 @@
         "Mio selector (offline)": {
           "method_id": "mio",
           "prediction_sha256": {
-            "20260917": "42f74906309c2cb4ca15c8236cd391e2f19a4830509854252fa33a5136fde791"
           },
           "splits": {
             "robust": {
@@ -3413,7 +3406,7 @@
         "Avengers": {
           "method_id": "avengers",
           "prediction_sha256": {
-            "20260917": "33ccae8c548d15f815072b0cbdbd8292004457728fbe0886eb9c34bea66e05b2"
           },
           "splits": {
             "robust": {
@@ -6652,7 +6645,7 @@
         "Avengers-Pro": {
           "method_id": "avengerspro",
           "prediction_sha256": {
-            "20260917": "b6dd7fb76865fd62644a3e31ac6bbc1ff269672f8565cf6a28b9eb60eb1bbfe6"
           },
           "splits": {
             "robust": {
@@ -9894,7 +9887,7 @@
         "Causal RM-Softmax": {
           "method_id": "causal_rm_softmax",
           "prediction_sha256": {
-            "20260917": "79422258eda8b38877ae58d7f0c6e6cc02fb1fe1d83954f9f8201e78c2cb00ca"
           },
           "splits": {
             "robust": {
@@ -13127,7 +13120,7 @@
         "Eagle": {
           "method_id": "eagle",
           "prediction_sha256": {
-            "20260917": "6641f769e2a56544282ae1130b738185bf79004b3093801448deb00e35199503"
           },
           "splits": {
             "robust": {
@@ -16372,7 +16365,7 @@
         "EmbedLLM-RouteJudge": {
           "method_id": "embedllm_routejudge",
           "prediction_sha256": {
-            "20260917": "fcccc4de6c53855b96b0a56e4eb5dee35f5df7760a6d24c77e20edccb194ce5f"
           },
           "splits": {
             "robust": {
@@ -19650,7 +19643,7 @@
         "GraphRouter": {
           "method_id": "graphrouter",
           "prediction_sha256": {
-            "20260917": "40b5194d34da4db7eb6350f1cac0415e8928856e7abba093219e085701ff447a"
           },
           "splits": {
             "robust": {
@@ -22883,7 +22876,7 @@
         "HybridLLM": {
           "method_id": "hybridllm",
           "prediction_sha256": {
-            "20260917": "cb5f481e4b760fe686119e8bc3cac73e57e1069e3c96ff7a9f540ba8d2a102dc"
           },
           "splits": {
             "robust": {
@@ -26122,7 +26115,7 @@
         "kNN": {
           "method_id": "knn",
           "prediction_sha256": {
-            "20260917": "0cbee52d58f530c8597589d72a43013905e8314e98e8c7ff9ab64907da92412f"
           },
           "splits": {
             "robust": {
@@ -29361,7 +29354,7 @@
         "MLP": {
           "method_id": "mlp",
           "prediction_sha256": {
-            "20260917": "336d9b39dd75feb0f113531c42265878f43947cd8bc32a3827400d2c881f0736"
           },
           "splits": {
             "robust": {
@@ -32639,7 +32632,7 @@
         "MIRT": {
           "method_id": "mirt",
           "prediction_sha256": {
-            "20260917": "9420fdff881c83cac40683b7d4d5025c0e984f47bcd61358d07088e75cc5e47b"
           },
           "splits": {
             "robust": {
@@ -35872,7 +35865,7 @@
         "RouteLLM-MF": {
           "method_id": "routellm_mf",
           "prediction_sha256": {
-            "20260917": "9be60b5bf315b25ae1f690711321683bccf10b7bece48aa33c31521c1abf70ca"
           },
           "splits": {
             "robust": {
@@ -39114,7 +39107,7 @@
         "SVM Router": {
           "method_id": "svm_router",
           "prediction_sha256": {
-            "20260917": "3464f7d976c7203de4bb4d286c50e5112490e8ab73d948a4f90650bb6c247385"
           },
           "splits": {
             "robust": {
@@ -42352,11 +42345,10 @@
         }
       },
       "split_audit": {
-        "original_evaluation_rows": 7188,
-        "development_rows": 500,
-        "excluded_evaluation_rows": 606,
-        "final_evaluation_rows": 6582,
-        "development_group_overlap": 0
       }
     },
     "multimodal": {
@@ -42366,9 +42358,9 @@
         "Mio selector (offline)": {
           "method_id": "mio",
           "prediction_sha256": {
-            "20260917": "0495e163cc72f238f051688dced8cdd4fc513d18d6927b9df1251d98844c67d5",
-            "20260918": "fd5f777ef128a9260c73a33ad277ebe22112b0cf124394d50853c7c505f96084",
-            "20260919": "10938b71e521f42cfa71acefc16f0e0ec84461ae22b51a81e9f5f42b424e70d4"
           },
           "splits": {
             "robust": {
@@ -46279,9 +46271,9 @@
         "Avengers": {
           "method_id": "avengers",
           "prediction_sha256": {
-            "20260917": "d138b48dec24dac0584aeda9422f8b3582705c3f4f503594b259536fd2e62d1e",
-            "20260918": "e764a1cf3a67a9158f708e07b8cc8022500f4750f6421a597f636dd1ec01dd0b",
-            "20260919": "f11c349ea4108c9e43ce05c90c8c7427de030b4e933817406bc55cb666bba464"
           },
           "splits": {
             "robust": {
@@ -50171,9 +50163,9 @@
         "Avengers-Pro": {
           "method_id": "avengerspro",
           "prediction_sha256": {
-            "20260917": "294fb104d04ba694c0b65720484cd7ac26480479f9dac3044c6f59afc4b08952",
-            "20260918": "ba6163472beab5c2d10eae105600b572abddae22ff5dd556d417c8750384e9f1",
-            "20260919": "b71b48cf56b3aa60182d6712378f604eddfd5462632ebadf5fcb4a5bf329f064"
           },
           "splits": {
             "robust": {
@@ -54066,9 +54058,9 @@
         "Causal RM-Softmax": {
           "method_id": "causal_rm_softmax",
           "prediction_sha256": {
-            "20260917": "58235a8e779574b75efa8b1ba659aaaa55e7f9cfb6c7b18c4c3cf9742f228063",
-            "20260918": "8a893141b43a3277e0fc9d0d6a321255a7a2fe7390daa519b8148cf31427afd5",
-            "20260919": "47357e6861cf0a646da48f3a781cd7de608d721a9dcd31894c77a0d067431740"
           },
           "splits": {
             "robust": {
@@ -57946,9 +57938,9 @@
         "Eagle": {
           "method_id": "eagle",
           "prediction_sha256": {
-            "20260917": "55ea20e9814765107b17b1ee812dc101077a9872b2c0e0c755a215545df42385",
-            "20260918": "55ea20e9814765107b17b1ee812dc101077a9872b2c0e0c755a215545df42385",
-            "20260919": "55ea20e9814765107b17b1ee812dc101077a9872b2c0e0c755a215545df42385"
           },
           "splits": {
             "robust": {
@@ -61832,9 +61824,9 @@
         "EmbedLLM-RouteJudge": {
           "method_id": "embedllm_routejudge",
           "prediction_sha256": {
-            "20260917": "928f7a7d88af3eaa898b8f11e08151fbd370fe93a5408f7fe8f3384009125fc3",
-            "20260918": "d3bf73f6acf6a76373ca644ad6340085fe9c56dea1421b36d2b223bf71b8ce2d",
-            "20260919": "3ca1423981b8be12803b8d7f46c63757349114ef4b7c0bc319dd71dd3cafc358"
           },
           "splits": {
             "robust": {
@@ -65724,9 +65716,9 @@
         "GraphRouter": {
           "method_id": "graphrouter",
           "prediction_sha256": {
-            "20260917": "27501649c5b32ced9806ebf5a4a6a9e93a6b28b56fb215839cfd5eced8e598d9",
-            "20260918": "2f9f78432164b6b82bdd7a7d09e35a2f2a83c523eca4fe48c340e3ba4f8e9658",
-            "20260919": "f71841a94c05e4103c13cda4eed89e184b7b9832b402e557f4c31c29cbeb0a52"
           },
           "splits": {
             "robust": {
@@ -69604,9 +69596,9 @@
         "HybridLLM": {
           "method_id": "hybridllm",
           "prediction_sha256": {
-            "20260917": "f7f2be4834dc56b5e2522c1b585ea0533ab2b76814152a078bc7fd001b39f097",
-            "20260918": "f7f2be4834dc56b5e2522c1b585ea0533ab2b76814152a078bc7fd001b39f097",
-            "20260919": "f7f2be4834dc56b5e2522c1b585ea0533ab2b76814152a078bc7fd001b39f097"
           },
           "splits": {
             "robust": {
@@ -73484,9 +73476,9 @@
         "kNN": {
           "method_id": "knn",
           "prediction_sha256": {
-            "20260917": "fa825cf4abdefd016b186eb3ca5874bce87c7f58baa3d460fb9450d71f36cd7a",
-            "20260918": "fa825cf4abdefd016b186eb3ca5874bce87c7f58baa3d460fb9450d71f36cd7a",
-            "20260919": "fa825cf4abdefd016b186eb3ca5874bce87c7f58baa3d460fb9450d71f36cd7a"
           },
           "splits": {
             "robust": {
@@ -77364,9 +77356,9 @@
         "MLP": {
           "method_id": "mlp",
           "prediction_sha256": {
-            "20260917": "960fd3e8d44ef13ca09a89ccf8c2215b9631a9c1fbdabcb0079520e16c998e1e",
-            "20260918": "0e899491b0e3b75cea8ddf0435f0f5e6613538fbd7eb00ee8cff8417c81ac1ed",
-            "20260919": "0a1fed68e51a646a513958f6b5a4d61a6afd7fe6b255901d9dcca0a72bfe36c7"
           },
           "splits": {
             "robust": {
@@ -81259,9 +81251,9 @@
         "MIRT": {
           "method_id": "mirt",
           "prediction_sha256": {
-            "20260917": "4741bafd50954f0f9c6e079f88166041ab1e1b5a715438b4cf574418cd121ed9",
-            "20260918": "6947410d98ea3360670ede2217d537d19f872055fb601ccf7c9267ccf6f8bf85",
-            "20260919": "927b6e1adcfe3d8a64cd7de9881ecfb0ff720f00cf75d7bf67b427f96008bf84"
           },
           "splits": {
             "robust": {
@@ -85139,9 +85131,9 @@
         "RouteLLM-MF": {
           "method_id": "routellm_mf",
           "prediction_sha256": {
-            "20260917": "76bbeef42e1eecba46a513baaf1c2340d44758b7a8160ff3d75aaa4977145df9",
-            "20260918": "76bbeef42e1eecba46a513baaf1c2340d44758b7a8160ff3d75aaa4977145df9",
-            "20260919": "76bbeef42e1eecba46a513baaf1c2340d44758b7a8160ff3d75aaa4977145df9"
           },
           "splits": {
             "robust": {
@@ -89022,9 +89014,9 @@
         "SVM Router": {
           "method_id": "svm_router",
           "prediction_sha256": {
-            "20260917": "ed3245fd11143ee8d12e15ca7084225bf1e9674029113c45e124be58aeeb5e51",
-            "20260918": "760f232c71bc7de11988c3ace31f36b510c6ee2ace218bc6c5fc9bf9ba58f173",
-            "20260919": "bdfb8fc94cef61fbc51dfe6e6a20c99ee3a8195881e646f1398300c942b4cf52"
           },
           "splits": {
             "robust": {
@@ -92901,11 +92893,10 @@
         }
       },
       "split_audit": {
-        "original_evaluation_rows": 7000,
-        "development_rows": 500,
-        "excluded_evaluation_rows": 500,
-        "final_evaluation_rows": 6500,
-        "development_group_overlap": 0
       },
       "feature_provenance": {
         "backend": "local_transformers_float32",
@@ -92939,7 +92930,7 @@
         "Mio selector (offline)": {
           "method_id": "mio",
           "prediction_sha256": {
-            "20260917": "d86e56ac54aba21f9863eb55ca763059a4b8a536628be28338575282938aeb3b"
           },
           "splits": {
             "robust": {
@@ -96205,7 +96196,7 @@
         "Avengers": {
           "method_id": "avengers",
           "prediction_sha256": {
-            "20260917": "fcd8b31ee2ee72646f2ce92615c187afafe5c211bbad687c2548c1f4b3299e7a"
           },
           "splits": {
             "robust": {
@@ -99465,7 +99456,7 @@
         "Avengers-Pro": {
           "method_id": "avengerspro",
           "prediction_sha256": {
-            "20260917": "df33ceb06775631501c05cb94e88678d1dbefda69efa49ed425135b96587e841"
           },
           "splits": {
             "robust": {
@@ -102737,7 +102728,7 @@
         "Causal RM-Softmax": {
           "method_id": "causal_rm_softmax",
           "prediction_sha256": {
-            "20260917": "5f32a2c48622299fac83beba10fce768e2f7f230a47a1581456015f8e44f9a06"
           },
           "splits": {
             "robust": {
@@ -105985,7 +105976,7 @@
         "Eagle": {
           "method_id": "eagle",
           "prediction_sha256": {
-            "20260917": "8757ffcb10bc0aaca23215d7c57a1d0c52e83109ef479c1f636363213b6cb7d6"
           },
           "splits": {
             "robust": {
@@ -109257,7 +109248,7 @@
         "EmbedLLM-RouteJudge": {
           "method_id": "embedllm_routejudge",
           "prediction_sha256": {
-            "20260917": "e0f5bce13e75893354af644612cf584ba0f90c62302e870acf7fd54cf701166a"
           },
           "splits": {
             "robust": {
@@ -112529,7 +112520,7 @@
         "GraphRouter": {
           "method_id": "graphrouter",
           "prediction_sha256": {
-            "20260917": "1aa4dde516ef3dd11632cf378ff1f3f37b8e8631fa7c672d05b07a756efb10a1"
           },
           "splits": {
             "robust": {
@@ -115804,7 +115795,7 @@
         "HybridLLM": {
           "method_id": "hybridllm",
           "prediction_sha256": {
-            "20260917": "b658f6bbd5f7cc37291c7065a8163b21796fd7f92c8637dfde02806a0b40458c"
           },
           "splits": {
             "robust": {
@@ -119067,7 +119058,7 @@
         "kNN": {
           "method_id": "knn",
           "prediction_sha256": {
-            "20260917": "3100b7dbe440a6c51280cd18598feb127c2aca332492573cfdd5c892fa32c514"
           },
           "splits": {
             "robust": {
@@ -122327,7 +122318,7 @@
         "MLP": {
           "method_id": "mlp",
           "prediction_sha256": {
-            "20260917": "8571bc321a4ce6516a87ea1ed1b0a37f823b5239147ececcdbc1f6fd27b89706"
           },
           "splits": {
             "robust": {
@@ -125599,7 +125590,7 @@
         "MIRT": {
           "method_id": "mirt",
           "prediction_sha256": {
-            "20260917": "6359fa117a93e85c6f79080343da6e90e801e193fa8d03b43437c9c601100784"
           },
           "splits": {
             "robust": {
@@ -128877,7 +128868,7 @@
         "RouteLLM-MF": {
           "method_id": "routellm_mf",
           "prediction_sha256": {
-            "20260917": "53214ad4dfa02c64110a0c1b0e022f778de2c748d2c5ba8b51c094ab08e7b80f"
           },
           "splits": {
             "robust": {
@@ -132137,7 +132128,7 @@
         "SVM Router": {
           "method_id": "svm_router",
           "prediction_sha256": {
-            "20260917": "3bce58280e389fde06d9e7de49d4bb8d21664d9f0d42aaa655c3dbd763d18b80"
           },
           "splits": {
             "robust": {
@@ -135384,11 +135375,10 @@
         }
       },
       "split_audit": {
-        "original_evaluation_rows": 6500,
-        "development_rows": 500,
-        "excluded_evaluation_rows": 600,
-        "final_evaluation_rows": 5900,
-        "development_group_overlap": 0
       }
     }
   }

 {
   "schema_version": "routestressbench-report-v3",
+  "generated_at": "2026-10-17T16:41:21.854016+00:00",
   "protocol": {
+    "data_revision": "v2.0.0",
     "ranked_implementations": 13,
     "baseline_implementations": 12,
     "budget_points": 100,
     "auc_axis": "common nominal per-request cost budget",
     "auc_envelope": "cumulative maximum router quality over increasing budgets",
     "fixed_model_references_in_router_curve": false,
+    "test_population": "official test split; all requests are scored",
+    "validation_test_source_group_overlap": 0,
     "multimodal_seeds": [
       20260917,
       20260918,
     "target_cost_estimator": "minimum realized mean cost on the pointwise mean curve",
     "target_reach_count": "number of individual runs reaching the target",
     "macro_std_identity": "descriptive sample standard deviation across six split scores",
+    "alias_audit_scope": "historical adapter audit; only 12 canonical baselines are ranked"
   },
   "macro": {
     "Mio selector (offline)": {
         "Mio selector (offline)": {
           "method_id": "mio",
           "prediction_sha256": {
+            "20260917": "1df6f5e58aa65cc7feb2e3589a88b451a6a80cd7de7f0712d34a78945615768e"
           },
           "splits": {
             "robust": {
         "Avengers": {
           "method_id": "avengers",
           "prediction_sha256": {
+            "20260917": "d323b2ca2a72cfe511680447c1b2e64cf4ecf7637be5d57130777a58cf048dcb"
           },
           "splits": {
             "robust": {
         "Avengers-Pro": {
           "method_id": "avengerspro",
           "prediction_sha256": {
+            "20260917": "65fd3fb6fad3f9ccd390cfa414a4c90157a5d88b2c9c8d22a769450de11eee10"
           },
           "splits": {
             "robust": {
         "Causal RM-Softmax": {
           "method_id": "causal_rm_softmax",
           "prediction_sha256": {
+            "20260917": "951e633198248a35738832bb27efedf2b9a7a5e42d1281798b7bd5cd7147c2f5"
           },
           "splits": {
             "robust": {
         "Eagle": {
           "method_id": "eagle",
           "prediction_sha256": {
+            "20260917": "85f0880c6de3c09fba3ba3fa996c91bb38c1fd1d5928ed55ee77814eff3bd598"
           },
           "splits": {
             "robust": {
         "EmbedLLM-RouteJudge": {
           "method_id": "embedllm_routejudge",
           "prediction_sha256": {
+            "20260917": "2397353f2de8f1dbb94faabf1debde09aa9cc5520d3ad670fd0042033215b9bc"
           },
           "splits": {
             "robust": {
         "GraphRouter": {
           "method_id": "graphrouter",
           "prediction_sha256": {
+            "20260917": "7b3b999f51e1202e146eb55b5137fb5b2942c31f881fe82dd5d46c2c1bd812cb"
           },
           "splits": {
             "robust": {
         "HybridLLM": {
           "method_id": "hybridllm",
           "prediction_sha256": {
+            "20260917": "7132f8e63711014f8dd3d5d49b3591baf865efc31d61a8961896c60b76b4cd21"
           },
           "splits": {
             "robust": {
         "kNN": {
           "method_id": "knn",
           "prediction_sha256": {
+            "20260917": "f5971af48d1f9fd8e0408c6025ecf74787b4163198648f5783a1d337b47a2899"
           },
           "splits": {
             "robust": {
         "MLP": {
           "method_id": "mlp",
           "prediction_sha256": {
+            "20260917": "8bf849d38a36f7e93b57d25edf0ba9039dbcf3c5c3bebd0a13271b4130993642"
           },
           "splits": {
             "robust": {
         "MIRT": {
           "method_id": "mirt",
           "prediction_sha256": {
+            "20260917": "a84a271053012fd92840d97a3d80c6ea9194f83ea4caf605a1265660f79c8468"
           },
           "splits": {
             "robust": {
         "RouteLLM-MF": {
           "method_id": "routellm_mf",
           "prediction_sha256": {
+            "20260917": "beb729ee903c61437f9e4e0efa7f12271ad51fc111987087f5d0a4c9c5ea9300"
           },
           "splits": {
             "robust": {
         "SVM Router": {
           "method_id": "svm_router",
           "prediction_sha256": {
+            "20260917": "b85f48e1fa13f3eb995eecfbbe6143e6eef73516b7cc0b5f3d9adfcaaaeca6a3"
           },
           "splits": {
             "robust": {
         }
       },
       "split_audit": {
+        "test_rows": 6582,
+        "validation_rows": 500,
+        "validation_test_query_overlap": 0,
+        "validation_test_source_group_overlap": 0
       }
     },
     "multimodal": {
         "Mio selector (offline)": {
           "method_id": "mio",
           "prediction_sha256": {
+            "20260917": "870bd8fd663cf6f817514736b0bd73ee7afabbabe34c445db97078c3002ab958",
+            "20260918": "8129221cf4b7039dd098b9ec4e7e093b322eb7e0e867446e48a9a5ddd6a43745",
+            "20260919": "44ae64729db23ff98929633272b88187606ce269eafebe5596e150c13e3f7eba"
           },
           "splits": {
             "robust": {
         "Avengers": {
           "method_id": "avengers",
           "prediction_sha256": {
+            "20260917": "56bb600409edecf4eda8c405ffade277da0e1d15fb324fe5b4a54c302c3e78cd",
+            "20260918": "3d765ede4e689e2b65e931cc9f70e3d67c7173724c27dd2b3728418261a890ba",
+            "20260919": "cba3e9fe713bee7375452c5344e9da3210183492a143f0f6dfb5e8c71b290af8"
           },
           "splits": {
             "robust": {
         "Avengers-Pro": {
           "method_id": "avengerspro",
           "prediction_sha256": {
+            "20260917": "3e759ac5882bdd02216d012017c71ff6f159fa2c8ac76fdad5d7142946dbe4b0",
+            "20260918": "051135680faadef0f6d742c27d5c805dc555743dca6265c6188855445eac5c46",
+            "20260919": "f0a5d65295937f1f9bbadaee07bb3b315af3a01e117bebdc6088bfc8dfd03f1d"
           },
           "splits": {
             "robust": {
         "Causal RM-Softmax": {
           "method_id": "causal_rm_softmax",
           "prediction_sha256": {
+            "20260917": "c5efff74f2d571f386a5a30f88fe99e3eba9742066a1cb5579e779c51d92093a",
+            "20260918": "5e9808dc1f2cff137e7d61480067e3be7b64b9ed5ac920fae957be29b28a0ea7",
+            "20260919": "5a880cf4aa9749f866e6142b6da6f40982dafdb04b86414104f9f9b71299dcca"
           },
           "splits": {
             "robust": {
         "Eagle": {
           "method_id": "eagle",
           "prediction_sha256": {
+            "20260917": "5b722e011383e4e6f59d4b16cef602af7c37fecd893fc95c1d15d753475e9b70",
+            "20260918": "5b722e011383e4e6f59d4b16cef602af7c37fecd893fc95c1d15d753475e9b70",
+            "20260919": "5b722e011383e4e6f59d4b16cef602af7c37fecd893fc95c1d15d753475e9b70"
           },
           "splits": {
             "robust": {
         "EmbedLLM-RouteJudge": {
           "method_id": "embedllm_routejudge",
           "prediction_sha256": {
+            "20260917": "55ba38330457440076f416374f0626bd0d57c1bf1da19c41477d81ae288b238f",
+            "20260918": "1b723a8cd05ea68f40e60480c2089a102f8afbab3364228963028327aafe3cab",
+            "20260919": "58c53b1053557a3c472f9b782a51cfa846fb099df0b59ca2a50e77fbe12466db"
           },
           "splits": {
             "robust": {
         "GraphRouter": {
           "method_id": "graphrouter",
           "prediction_sha256": {
+            "20260917": "dc796cf3080b0c1f467e8fb1a453f10550c50a573d3465049dc2d7b92025a0a7",
+            "20260918": "5d87d4b983842f20818c3e78711e1c372104cc897b9ce3311b050bbf942bcf03",
+            "20260919": "98389eea94b547b5e2333fecbdfb1349684a123c906638c03f93271902c145d8"
           },
           "splits": {
             "robust": {
         "HybridLLM": {
           "method_id": "hybridllm",
           "prediction_sha256": {
+            "20260917": "bfdfb93ce851e5935ce7521b6e34e9d9aeb1d5d70cc712294ab4d8e9113f455f",
+            "20260918": "bfdfb93ce851e5935ce7521b6e34e9d9aeb1d5d70cc712294ab4d8e9113f455f",
+            "20260919": "bfdfb93ce851e5935ce7521b6e34e9d9aeb1d5d70cc712294ab4d8e9113f455f"
           },
           "splits": {
             "robust": {
         "kNN": {
           "method_id": "knn",
           "prediction_sha256": {
+            "20260917": "f028ff9bb2d27d0f39897da1a997bf3f0fa9c62f76d2e7e5b67fa965c4119aba",
+            "20260918": "f028ff9bb2d27d0f39897da1a997bf3f0fa9c62f76d2e7e5b67fa965c4119aba",
+            "20260919": "f028ff9bb2d27d0f39897da1a997bf3f0fa9c62f76d2e7e5b67fa965c4119aba"
           },
           "splits": {
             "robust": {
         "MLP": {
           "method_id": "mlp",
           "prediction_sha256": {
+            "20260917": "5b5f2d248ea1876520a7022ed50c88da09a7ce140c8ca72c2052b33a6986f68b",
+            "20260918": "21dc30beaa0fd6adf14c11d05c7012d645eab8ba6b0eb47aa55ac5439432b1ed",
+            "20260919": "9d0cbb5796665fe0d4b66a2cda2de254c6fa0a34a8cd3b841c3716f5adda82c8"
           },
           "splits": {
             "robust": {
         "MIRT": {
           "method_id": "mirt",
           "prediction_sha256": {
+            "20260917": "1d3bb466a75a9884f9f77b9989866113b434fc3075c397e4680caa2912bc9500",
+            "20260918": "e07a5ca6e359768f243e2fc20f24022d259c118b5d379d5917032663cdca3465",
+            "20260919": "2b288a46f3ca23e462c33dcee625cb67603c17a605c19cfa2dc30a4cca3eaab2"
           },
           "splits": {
             "robust": {
         "RouteLLM-MF": {
           "method_id": "routellm_mf",
           "prediction_sha256": {
+            "20260917": "d2ecb163d808fe4481fa50ebb926937d13bab58eaf49492edf3309a8a5eef252",
+            "20260918": "d2ecb163d808fe4481fa50ebb926937d13bab58eaf49492edf3309a8a5eef252",
+            "20260919": "d2ecb163d808fe4481fa50ebb926937d13bab58eaf49492edf3309a8a5eef252"
           },
           "splits": {
             "robust": {
         "SVM Router": {
           "method_id": "svm_router",
           "prediction_sha256": {
+            "20260917": "741569e6b9c01ee8ff16a789d8d5b01477c6274407d255814554a24e81d024c3",
+            "20260918": "980a68980e8741c9c2ecdb322fc3c80dfa7612847dbb215a5719b5dc926b10f8",
+            "20260919": "8108f519ae9bdd81ef1bfd9bde72a9cf0ed81ab0cfa863b47288ff9398b63cf5"
           },
           "splits": {
             "robust": {
         }
       },
       "split_audit": {
+        "test_rows": 6500,
+        "validation_rows": 500,
+        "validation_test_query_overlap": 0,
+        "validation_test_source_group_overlap": 0
       },
       "feature_provenance": {
         "backend": "local_transformers_float32",
         "Mio selector (offline)": {
           "method_id": "mio",
           "prediction_sha256": {
+            "20260917": "861d3fd5110b55169ffcf4e1d6d099ebb982207320e4253a487a8f1f193c51be"
           },
           "splits": {
             "robust": {
         "Avengers": {
           "method_id": "avengers",
           "prediction_sha256": {
+            "20260917": "fac73ad49bb6eb4365d50458e006b650dbd9aef782148e683839767d29db613f"
           },
           "splits": {
             "robust": {
         "Avengers-Pro": {
           "method_id": "avengerspro",
           "prediction_sha256": {
+            "20260917": "c6d4c99f3936800022af5bdcc345f496cef3f27046411a539002e381172b9144"
           },
           "splits": {
             "robust": {
         "Causal RM-Softmax": {
           "method_id": "causal_rm_softmax",
           "prediction_sha256": {
+            "20260917": "db1da64aaeeb5272adc963158870c04c5029bf4d30bc762e9ffaa856faa78f78"
           },
           "splits": {
             "robust": {
         "Eagle": {
           "method_id": "eagle",
           "prediction_sha256": {
+            "20260917": "5fbe1650e1728e89444f2a408bf763e50dc00def59dc223fd4b5032c156544a5"
           },
           "splits": {
             "robust": {
         "EmbedLLM-RouteJudge": {
           "method_id": "embedllm_routejudge",
           "prediction_sha256": {
+            "20260917": "9798166e4d0d2949379a1bdecb96dce0c02e15aa08a116e0c68a9bfdb040dae2"
           },
           "splits": {
             "robust": {
         "GraphRouter": {
           "method_id": "graphrouter",
           "prediction_sha256": {
+            "20260917": "e06d7fdfb162e6341b0462f82570764c28d59b28b1373af15dce9031ff7a01be"
           },
           "splits": {
             "robust": {
         "HybridLLM": {
           "method_id": "hybridllm",
           "prediction_sha256": {
+            "20260917": "78137fcbc20154551a8a42079ddfcb30fe6a189f9c3c4befe5ac04cd0905b870"
           },
           "splits": {
             "robust": {
         "kNN": {
           "method_id": "knn",
           "prediction_sha256": {
+            "20260917": "ab4909a7e095dd1858839794271db91ffea805e015bdcecf199ef354259f917f"
           },
           "splits": {
             "robust": {
         "MLP": {
           "method_id": "mlp",
           "prediction_sha256": {
+            "20260917": "44bd905bc2e01d958d9e70e660c31ab1b1aee91b5dc65d3fbd02d61dfb398a22"
           },
           "splits": {
             "robust": {
         "MIRT": {
           "method_id": "mirt",
           "prediction_sha256": {
+            "20260917": "949034a807451176f04874305d7afc2ddaf38f7df8d97b79256235b9aa0fa62b"
           },
           "splits": {
             "robust": {
         "RouteLLM-MF": {
           "method_id": "routellm_mf",
           "prediction_sha256": {
+            "20260917": "82a7228bce9282e30f928648312768d2f5b20b0ba4c487992263902c5046e536"
           },
           "splits": {
             "robust": {
         "SVM Router": {
           "method_id": "svm_router",
           "prediction_sha256": {
+            "20260917": "a14af06dd1213fccd6f5c9cd68ec24392c119a4766140018510102c2d3ee298b"
           },
           "splits": {
             "robust": {
         }
       },
       "split_audit": {
+        "test_rows": 5900,
+        "validation_rows": 500,
+        "validation_test_query_overlap": 0,
+        "validation_test_source_group_overlap": 0
       }
     }
   }

report/predictions/chinese/seed20260917/avengers.json CHANGED Viewed

@@ -3467,13 +3467,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "fcd8b31ee2ee72646f2ce92615c187afafe5c211bbad687c2548c1f4b3299e7a",
   "seed": 20260917,
   "method_id": "avengers",
   "config": {
@@ -3520,5 +3519,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "fac73ad49bb6eb4365d50458e006b650dbd9aef782148e683839767d29db613f",
   "seed": 20260917,
   "method_id": "avengers",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/avengers.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fcd8b31ee2ee72646f2ce92615c187afafe5c211bbad687c2548c1f4b3299e7a
-size 142401

 version https://git-lfs.github.com/spec/v1
+oid sha256:fac73ad49bb6eb4365d50458e006b650dbd9aef782148e683839767d29db613f
+size 131564

report/predictions/chinese/seed20260917/avengerspro.json CHANGED Viewed

@@ -3699,13 +3699,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "df33ceb06775631501c05cb94e88678d1dbefda69efa49ed425135b96587e841",
   "seed": 20260917,
   "method_id": "avengerspro",
   "config": {
@@ -3753,5 +3752,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "c6d4c99f3936800022af5bdcc345f496cef3f27046411a539002e381172b9144",
   "seed": 20260917,
   "method_id": "avengerspro",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/avengerspro.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:df33ceb06775631501c05cb94e88678d1dbefda69efa49ed425135b96587e841
-size 266780

 version https://git-lfs.github.com/spec/v1
+oid sha256:c6d4c99f3936800022af5bdcc345f496cef3f27046411a539002e381172b9144
+size 248661

report/predictions/chinese/seed20260917/causal_rm_softmax.json CHANGED Viewed

@@ -3339,13 +3339,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "5f32a2c48622299fac83beba10fce768e2f7f230a47a1581456015f8e44f9a06",
   "seed": 20260917,
   "method_id": "causal_rm_softmax",
   "config": {
@@ -3392,5 +3391,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "db1da64aaeeb5272adc963158870c04c5029bf4d30bc762e9ffaa856faa78f78",
   "seed": 20260917,
   "method_id": "causal_rm_softmax",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/causal_rm_softmax.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5f32a2c48622299fac83beba10fce768e2f7f230a47a1581456015f8e44f9a06
-size 719287

 version https://git-lfs.github.com/spec/v1
+oid sha256:db1da64aaeeb5272adc963158870c04c5029bf4d30bc762e9ffaa856faa78f78
+size 655113

report/predictions/chinese/seed20260917/eagle.json CHANGED Viewed

@@ -3835,13 +3835,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "8757ffcb10bc0aaca23215d7c57a1d0c52e83109ef479c1f636363213b6cb7d6",
   "seed": 20260917,
   "method_id": "eagle",
   "config": {
@@ -3889,5 +3888,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "5fbe1650e1728e89444f2a408bf763e50dc00def59dc223fd4b5032c156544a5",
   "seed": 20260917,
   "method_id": "eagle",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/eagle.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8757ffcb10bc0aaca23215d7c57a1d0c52e83109ef479c1f636363213b6cb7d6
-size 776117

 version https://git-lfs.github.com/spec/v1
+oid sha256:5fbe1650e1728e89444f2a408bf763e50dc00def59dc223fd4b5032c156544a5
+size 705901

report/predictions/chinese/seed20260917/embedllm_routejudge.json CHANGED Viewed

@@ -3263,13 +3263,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "e0f5bce13e75893354af644612cf584ba0f90c62302e870acf7fd54cf701166a",
   "seed": 20260917,
   "method_id": "embedllm_routejudge",
   "config": {
@@ -3316,5 +3315,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "9798166e4d0d2949379a1bdecb96dce0c02e15aa08a116e0c68a9bfdb040dae2",
   "seed": 20260917,
   "method_id": "embedllm_routejudge",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/embedllm_routejudge.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e0f5bce13e75893354af644612cf584ba0f90c62302e870acf7fd54cf701166a
-size 321990

 version https://git-lfs.github.com/spec/v1
+oid sha256:9798166e4d0d2949379a1bdecb96dce0c02e15aa08a116e0c68a9bfdb040dae2
+size 294685

report/predictions/chinese/seed20260917/graphrouter.json CHANGED Viewed

@@ -3323,13 +3323,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "1aa4dde516ef3dd11632cf378ff1f3f37b8e8631fa7c672d05b07a756efb10a1",
   "seed": 20260917,
   "method_id": "graphrouter",
   "config": {
@@ -3376,5 +3375,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "e06d7fdfb162e6341b0462f82570764c28d59b28b1373af15dce9031ff7a01be",
   "seed": 20260917,
   "method_id": "graphrouter",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/graphrouter.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1aa4dde516ef3dd11632cf378ff1f3f37b8e8631fa7c672d05b07a756efb10a1
-size 762504

 version https://git-lfs.github.com/spec/v1
+oid sha256:e06d7fdfb162e6341b0462f82570764c28d59b28b1373af15dce9031ff7a01be
+size 694463

report/predictions/chinese/seed20260917/hybridllm.json CHANGED Viewed

@@ -4131,13 +4131,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "b658f6bbd5f7cc37291c7065a8163b21796fd7f92c8637dfde02806a0b40458c",
   "seed": 20260917,
   "method_id": "hybridllm",
   "config": {
@@ -4184,5 +4183,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "78137fcbc20154551a8a42079ddfcb30fe6a189f9c3c4befe5ac04cd0905b870",
   "seed": 20260917,
   "method_id": "hybridllm",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/hybridllm.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b658f6bbd5f7cc37291c7065a8163b21796fd7f92c8637dfde02806a0b40458c
-size 767849

 version https://git-lfs.github.com/spec/v1
+oid sha256:78137fcbc20154551a8a42079ddfcb30fe6a189f9c3c4befe5ac04cd0905b870
+size 698950

report/predictions/chinese/seed20260917/knn.json CHANGED Viewed

@@ -4067,13 +4067,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "3100b7dbe440a6c51280cd18598feb127c2aca332492573cfdd5c892fa32c514",
   "seed": 20260917,
   "method_id": "knn",
   "config": {
@@ -4120,5 +4119,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "ab4909a7e095dd1858839794271db91ffea805e015bdcecf199ef354259f917f",
   "seed": 20260917,
   "method_id": "knn",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/knn.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3100b7dbe440a6c51280cd18598feb127c2aca332492573cfdd5c892fa32c514
-size 776732

 version https://git-lfs.github.com/spec/v1
+oid sha256:ab4909a7e095dd1858839794271db91ffea805e015bdcecf199ef354259f917f
+size 706605

report/predictions/chinese/seed20260917/mio.json CHANGED Viewed

@@ -3875,14 +3875,14 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "d86e56ac54aba21f9863eb55ca763059a4b8a536628be28338575282938aeb3b",
   "seed": 20260917,
   "method_id": "mio",
-  "training_reproduced": false
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "861d3fd5110b55169ffcf4e1d6d099ebb982207320e4253a487a8f1f193c51be",
   "seed": 20260917,
   "method_id": "mio",
+  "training_reproduced": false,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/mio.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d86e56ac54aba21f9863eb55ca763059a4b8a536628be28338575282938aeb3b
-size 786641

 version https://git-lfs.github.com/spec/v1
+oid sha256:861d3fd5110b55169ffcf4e1d6d099ebb982207320e4253a487a8f1f193c51be
+size 714312

report/predictions/chinese/seed20260917/mirt.json CHANGED Viewed

@@ -3463,13 +3463,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "6359fa117a93e85c6f79080343da6e90e801e193fa8d03b43437c9c601100784",
   "seed": 20260917,
   "method_id": "mirt",
   "config": {
@@ -3516,5 +3515,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "949034a807451176f04874305d7afc2ddaf38f7df8d97b79256235b9aa0fa62b",
   "seed": 20260917,
   "method_id": "mirt",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/mirt.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6359fa117a93e85c6f79080343da6e90e801e193fa8d03b43437c9c601100784
-size 736936

 version https://git-lfs.github.com/spec/v1
+oid sha256:949034a807451176f04874305d7afc2ddaf38f7df8d97b79256235b9aa0fa62b
+size 670605

report/predictions/chinese/seed20260917/mlp.json CHANGED Viewed

@@ -3803,13 +3803,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "8571bc321a4ce6516a87ea1ed1b0a37f823b5239147ececcdbc1f6fd27b89706",
   "seed": 20260917,
   "method_id": "mlp",
   "config": {
@@ -3856,5 +3855,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "44bd905bc2e01d958d9e70e660c31ab1b1aee91b5dc65d3fbd02d61dfb398a22",
   "seed": 20260917,
   "method_id": "mlp",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/mlp.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8571bc321a4ce6516a87ea1ed1b0a37f823b5239147ececcdbc1f6fd27b89706
-size 781325

 version https://git-lfs.github.com/spec/v1
+oid sha256:44bd905bc2e01d958d9e70e660c31ab1b1aee91b5dc65d3fbd02d61dfb398a22
+size 710721

report/predictions/chinese/seed20260917/routellm_mf.json CHANGED Viewed

@@ -4039,13 +4039,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "53214ad4dfa02c64110a0c1b0e022f778de2c748d2c5ba8b51c094ab08e7b80f",
   "seed": 20260917,
   "method_id": "routellm_mf",
   "config": {
@@ -4092,5 +4091,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "82a7228bce9282e30f928648312768d2f5b20b0ba4c487992263902c5046e536",
   "seed": 20260917,
   "method_id": "routellm_mf",
   "config": {
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/routellm_mf.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:53214ad4dfa02c64110a0c1b0e022f778de2c748d2c5ba8b51c094ab08e7b80f
-size 768060

 version https://git-lfs.github.com/spec/v1
+oid sha256:82a7228bce9282e30f928648312768d2f5b20b0ba4c487992263902c5046e536
+size 699205

report/predictions/chinese/seed20260917/svm_router.json CHANGED Viewed

@@ -3783,13 +3783,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 6500,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 600,
-    "final_evaluation_rows": 5900,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "3bce58280e389fde06d9e7de49d4bb8d21664d9f0d42aaa655c3dbd763d18b80",
   "seed": 20260917,
   "method_id": "svm_router",
   "config": {
@@ -3838,5 +3837,6 @@
     "ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.": 30
   },
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 5900,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "a14af06dd1213fccd6f5c9cd68ec24392c119a4766140018510102c2d3ee298b",
   "seed": 20260917,
   "method_id": "svm_router",
   "config": {
     "ConvergenceWarning: Liblinear failed to converge, increase the number of iterations.": 30
   },
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/chinese/seed20260917/svm_router.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3bce58280e389fde06d9e7de49d4bb8d21664d9f0d42aaa655c3dbd763d18b80
-size 745694

 version https://git-lfs.github.com/spec/v1
+oid sha256:a14af06dd1213fccd6f5c9cd68ec24392c119a4766140018510102c2d3ee298b
+size 674338

report/predictions/multimodal/seed20260917/avengers.json CHANGED Viewed

@@ -3956,13 +3956,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 7000,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 500,
-    "final_evaluation_rows": 6500,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "d138b48dec24dac0584aeda9422f8b3582705c3f4f503594b259536fd2e62d1e",
   "seed": 20260917,
   "method_id": "avengers",
   "feature_dimension": 1024,
@@ -4034,5 +4033,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 6500,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "56bb600409edecf4eda8c405ffade277da0e1d15fb324fe5b4a54c302c3e78cd",
   "seed": 20260917,
   "method_id": "avengers",
   "feature_dimension": 1024,
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/multimodal/seed20260917/avengers.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d138b48dec24dac0584aeda9422f8b3582705c3f4f503594b259536fd2e62d1e
-size 154891

 version https://git-lfs.github.com/spec/v1
+oid sha256:56bb600409edecf4eda8c405ffade277da0e1d15fb324fe5b4a54c302c3e78cd
+size 147504

report/predictions/multimodal/seed20260917/avengerspro.json CHANGED Viewed

@@ -4308,13 +4308,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 7000,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 500,
-    "final_evaluation_rows": 6500,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "294fb104d04ba694c0b65720484cd7ac26480479f9dac3044c6f59afc4b08952",
   "seed": 20260917,
   "method_id": "avengerspro",
   "feature_dimension": 1024,
@@ -4387,5 +4386,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 6500,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "3e759ac5882bdd02216d012017c71ff6f159fa2c8ac76fdad5d7142946dbe4b0",
   "seed": 20260917,
   "method_id": "avengerspro",
   "feature_dimension": 1024,
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/multimodal/seed20260917/avengerspro.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:294fb104d04ba694c0b65720484cd7ac26480479f9dac3044c6f59afc4b08952
-size 635725

 version https://git-lfs.github.com/spec/v1
+oid sha256:3e759ac5882bdd02216d012017c71ff6f159fa2c8ac76fdad5d7142946dbe4b0
+size 599860

report/predictions/multimodal/seed20260917/causal_rm_softmax.json CHANGED Viewed

@@ -3876,13 +3876,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 7000,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 500,
-    "final_evaluation_rows": 6500,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "58235a8e779574b75efa8b1ba659aaaa55e7f9cfb6c7b18c4c3cf9742f228063",
   "seed": 20260917,
   "method_id": "causal_rm_softmax",
   "feature_dimension": 1024,
@@ -3954,5 +3953,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 6500,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "c5efff74f2d571f386a5a30f88fe99e3eba9742066a1cb5579e779c51d92093a",
   "seed": 20260917,
   "method_id": "causal_rm_softmax",
   "feature_dimension": 1024,
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/multimodal/seed20260917/causal_rm_softmax.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:58235a8e779574b75efa8b1ba659aaaa55e7f9cfb6c7b18c4c3cf9742f228063
-size 3275540

 version https://git-lfs.github.com/spec/v1
+oid sha256:c5efff74f2d571f386a5a30f88fe99e3eba9742066a1cb5579e779c51d92093a
+size 3043127

report/predictions/multimodal/seed20260917/eagle.json CHANGED Viewed

@@ -4472,13 +4472,12 @@
     }
   },
   "split_audit": {
-    "original_evaluation_rows": 7000,
-    "development_rows": 500,
-    "excluded_evaluation_rows": 500,
-    "final_evaluation_rows": 6500,
-    "development_group_overlap": 0
   },
-  "prediction_sha256": "55ea20e9814765107b17b1ee812dc101077a9872b2c0e0c755a215545df42385",
   "seed": 20260917,
   "method_id": "eagle",
   "feature_dimension": 1024,
@@ -4551,5 +4550,6 @@
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
-  "training_reproduced": true
 }

     }
   },
   "split_audit": {
+    "test_rows": 6500,
+    "validation_rows": 500,
+    "validation_test_query_overlap": 0,
+    "validation_test_source_group_overlap": 0
   },
+  "prediction_sha256": "5b722e011383e4e6f59d4b16cef602af7c37fecd893fc95c1d15d753475e9b70",
   "seed": 20260917,
   "method_id": "eagle",
   "feature_dimension": 1024,
   },
   "fitting_warnings": {},
   "predicted_quality_clipped_to_unit_interval": true,
+  "training_reproduced": true,
+  "dataset_schema": "robust-route-bench-v2"
 }

report/predictions/multimodal/seed20260917/eagle.npz CHANGED Viewed

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:55ea20e9814765107b17b1ee812dc101077a9872b2c0e0c755a215545df42385
-size 2381404

 version https://git-lfs.github.com/spec/v1
+oid sha256:5b722e011383e4e6f59d4b16cef602af7c37fecd893fc95c1d15d753475e9b70
+size 2212904