Research

Evaluates Vision-Language Models on atomic world modeling capabilities across perception (spatial, temporal, motion) and prediction (mechanistic simulation, transitive/compositional inference) tasks. It probes whether VLMs possess internal representations of physical causality, dynamics, and multi-step reasoning comparable to human intuition. Use when the user wants to benchmark on WM-ABench, or asks about evaluating this task. Reports accuracy.

Probes instruction-following capability on complex, real-world prompts across diverse domains like coding, math, reasoning, and formatting. It measures how well models handle demanding, multi-step tasks compared to baselines through blind pairwise human comparison. Use when the user wants to benchmark on WizardEval, or asks about evaluating this task. Reports win_rate.

Evaluates human activity recognition (HAR) performance using only wireless Channel State Information (CSI) signals under varying action segmentation windows (1s, 2s, 3s). It probes the robustness of classification models in privacy-preserving environments where visual data is occluded or unavailable during testing. Use when the user wants to benchmark on WiVi, or asks about evaluating this task. Reports OA.

Evaluates a robot's ability to generalize visuomotor planning to unseen visual signals, referring expressions, and spatial relationships during cube-picking and placing tasks. It probes semantic generalization and robustness to distribution shifts in embodied instruction following. Use when the user wants to benchmark on WISER Benchmark, or asks about evaluating this task. Reports Success.

Evaluates Open Information Extraction systems on their ability to accurately extract relational tuples from text. It probes token-level precision and recall by matching predicted arguments and relations against a fine-grained, manually annotated gold standard. Use when the user wants to benchmark on WiRe57, or asks about evaluating this task. Reports token-weighted F1.

Evaluates whether language models disproportionately associate harmful stereotypes with marginalized groups (Jewish people or LGBTQ+ subgroups) compared to non-target groups. It also assesses the quality and reliability of automated versus human annotation for constructing community-sourced fairness benchmarks. Use when the user wants to benchmark on WinoSemitism, WinoQueer, or asks about evaluating this task. Reports WinoSem. Score.

Evaluates anti-LGBTQ+ bias in language models by measuring their tendency to prefer stereotypical completions over counterfactual ones when prompted with identity-specific contexts. Use when the user wants to benchmark on WinoQueer, or asks about evaluating this task. Reports bias score.

Probes vision-language models' ability to understand visio-linguistic compositionality and word order sensitivity. The task requires matching images to captions where identical words are rearranged to change the described scene, testing structural grounding rather than lexical overlap. Use when the user wants to benchmark on Winoground, or asks about evaluating this task. Reports image-caption score.

This benchmark probes systematic gender bias in coreference resolution systems by measuring how often models resolve gendered pronouns to occupations differently based solely on pronoun gender. It evaluates whether models reinforce real-world occupational gender disparities and how performance degrades on counter-stereotypical ('gotcha') examples. Use when the user wants to benchmark on Winogender schemas, or asks about evaluating this task. Reports bias_score.

Evaluates gender bias in coreference resolution systems by measuring performance disparity between pro-stereotypical and anti-stereotypical sentences. It probes whether models rely on gender stereotypes when resolving coreferences in challenging, Winograd-style contexts. Use when the user wants to benchmark on WinoBias, or asks about evaluating this task. Reports F1.

Evaluates large language models on comprehensive medical reasoning, clinical calculation, and general cognitive capabilities. It probes domain-specific knowledge application, diagnostic reasoning, and complex problem-solving in real-world clinical and academic settings. Use when the user wants to benchmark on MedCalc, MedReMCQ, CMMLU, MATH-500, MedQA-USMLE, MedMCQA, PubMedQA, or asks about evaluating this task. Reports accuracy.

Quantifies the intrinsic predictability of time series data by measuring window-wise pattern complexity in the frequency domain. It establishes a data-driven performance lower bound for forecasting models and identifies whether standard benchmarks have reached saturation. Use when the user has predictions and gold and needs to compute window-wise complexity.

Evaluates the calibration, sharpness, and accuracy of probabilistic wind power forecasts under different ensemble post-processing strategies (raw, weather-only, power-only, and joint weather-power post-processing). It probes whether correcting biases at the weather stage alone is sufficient, or if direct post-processing of the final power ensemble is required to handle non-linear power curve biases. Use when the user wants to benchmark on Benchmark Data, Swedish Data Set, or asks about evalua...

Evaluates the ability of models to correctly identify the language of monolingual text paragraphs. It probes language identification capabilities across a wide range of languages (235) with balanced representation. Use when the user wants to benchmark on WiLI-2018, or asks about evaluating this task. Reports F1.

This benchmark evaluates multimodal large language models' ability to perform multi-step, context-sensitive reasoning over symbolic musical notation. It probes capabilities in harmonic analysis, rhythmic interpretation, structural form recognition, and expressive markings through multiple-choice questions derived from real-world compositions and forum queries. Use when the user wants to benchmark on WildScore, or asks about evaluating this task. Reports accuracy.

Evaluates language models' scientific reasoning capabilities by testing their ability to answer domain-specific multiple-choice questions derived from peer-reviewed literature and established scientific benchmarks. Use when the user wants to benchmark on WildSci-Val, GPQA-Aug, SuperGPQA, MMLU-Pro, or asks about evaluating this task. Reports accuracy.

Probes the capability of models to perform semantic segmentation on unstructured, large-scale natural environments using both 2D images and 3D LiDAR point clouds. It evaluates robustness to semantic ambiguity, clutter, and temporal environmental shifts in outdoor traversals. Use when the user wants to benchmark on WildScenes, or asks about evaluating this task. Reports mIoU.

Evaluates machine learning models' robustness to real-world distribution shifts, specifically domain generalization and subpopulation shifts. It measures how much model performance degrades when tested on out-of-distribution (OOD) data compared to in-distribution (ID) data, highlighting gaps in generalization for real-world deployment. Use when the user wants to benchmark on WILDS, or asks about evaluating this task. Reports ID and OOD performance.

Evaluates novel view synthesis and motion mask estimation in dynamic environments where both camera and objects move. It probes a model's ability to remove transient objects, complete occluded backgrounds, and preserve scene geometry from sparse input views without 3D supervision or ground-truth poses. Use when the user wants to benchmark on D-RE10K-Mask, D-RE10K-iPhone, or asks about evaluating this task. Reports PSNR.

Evaluates the safety and robustness of language models against adversarial jailbreak attacks. It probes whether models can correctly refuse harmful requests while avoiding over-refusal on benign prompts, specifically under stealthy, adversarially composed prompts. Use when the user wants to benchmark on WILDJAILBREAK, or asks about evaluating this task. Reports Attack success rate (ASR).

This evaluation protocol assesses the safety moderation capabilities of LLMs and dedicated moderation models. It probes their ability to detect harmful content in user prompts, classify harmful or safe model responses, and identify whether a model appropriately refuses unsafe requests across multiple risk categories. Use when the user wants to benchmark on ToxicChat, OpenAI Mod, AegisSafetyTest, SimpleSafetyTests, Harmbench Prompt, Harmbench Resp, BeaverTails, SafeRLHF, XSTest-Resp, WildGuard...

This benchmark evaluates a model's ability to forecast the final spatial extent of a wildfire using multi-day spatio-temporal environmental and dynamic features. It probes the model's capacity to capture complex temporal dependencies and spatial patterns in binary segmentation tasks under significant class imbalance. Use when the user wants to benchmark on Mediterranean Wildfire Dataset (2006-2022), or asks about evaluating this task. Reports Dice Score.

This benchmark evaluates automatic speech recognition (ASR) capabilities on Mandarin speech produced by elderly individuals. It probes a model's robustness to real-world acoustic degradation, articulation variability, tremors, and diverse accent strengths under uncontrolled recording conditions. Use when the user wants to benchmark on WildElder, or asks about evaluating this task. Reports Word Error Rate (WER).

Evaluates open-vocabulary monocular 3D object detection across diverse real-world and synthetic scenes. It probes the model's ability to localize and regress 3D bounding boxes using text or geometric prompts, measuring generalization to unseen categories and datasets with and without depth cues. Use when the user wants to benchmark on WildDet3D-Bench, Omni3D, Argoverse 2, ScanNet, Stereo4D, or asks about evaluating this task. Reports AP_3D.