Smart search ranks by meaning as well as keywords (one row per story, last 45 days).
AllenAI's BenchMIRT blog post examines what LLM benchmarks actually measure and their reliability.
AllenAI published a Hugging Face blog post introducing BenchMIRT, which investigates what large language model benchmarks actually measure. No article text is available, so specific findings, methods, or benchmark scores cannot be extracted. The work appears to target benchmark validity, a live concern for model evaluation and comparison.
Version 0.34 of Simon Willison's llm CLI adds response-duration metrics to log output, plus bug fixes and faster log querying.
The open-source llm command-line tool for interacting with large language models released version 0.34. The headline change adds response duration in milliseconds and human-readable form to llm logs --usage Markdown output, plus a new duration_ms field in llm logs --short. The release includes several contributed bug fixes and a significant performance improvement to llm logs, alongside the related llm-openrouter 0.7.1 release.
Audit of eight cybersecurity LLM benchmarks shows evaluation pipeline choices can swing scores by over 80 points and reshuffle most model rankings.
Researchers modeled eight cybersecurity benchmarks as configurable measurement pipelines and audited 10 proprietary, open-weight, and cybersecurity-specialized LLMs. They identified 15 systematic failure modes and showed a single pipeline choice can change a model's score by more than 80 percentage points and alter rankings; semantically similar task pairs rank the same models differently. Under a standardized harness, nine of 10 models shifted at least three ranks on at least one benchmark, motivating pipeline-aware auditing for reliable model evaluation.
CS-Guard benchmark shows LLM code-generation guardrails fail widely, with ~50% jailbreak ASR text-to-code and up to 100% code-to-code.
Researchers introduce CS-Guard, the first systematic benchmark for evaluating LLM guardrails for code generation security, covering text-to-code (1,000 malware-generation prompts, 7 jailbreak attacks, and a novel fictional scenario attack) and code-to-code (331 prompts across infilling, completion, and translation). They evaluate 9 guardrails across seven LLMs, finding average jailbreak attack success rates around 50% for text-to-code and 14.4% to nearly 100% for code-to-code. The fictional scenario attack achieves ASR close to 100% across many guardrails, raising reliability concerns for real-world software development. The benchmark and data are released publicly.
MetroLLM-Bench is a 955-case benchmark testing language models as transit kiosk tool-calling runtimes across six real metro systems.
The benchmark covers 37-414-station metro systems and eleven task categories including routing, fare calculation, disruptions, accessibility, and adversarial input, with 14 deterministic and 8 semantic scoring components. Of 26 models from six vendors, a PEFT-tuned 4B Qwen 3.5 student scored 91.3 on Tier 1, exceeding GPT-5.6 (90.6/90.0), while Muse Glimmer 30B led the composite ranking. A deterministic rule-based baseline reached 84.6, and PEFT gains over base models shrank from +7.03 points at 2B to -0.91 at 27B.
TIER benchmark shows LLM safety behaviors shift gradually across threat implicitness levels, with jailbreaks exposing the largest robustness gaps.
The TIER benchmark evaluates LLM safety behaviors across four risk domains and four threat levels, from explicit harmful requests to sophisticated jailbreaks, using a six-label behavior scale and two independent LLM judges. Experiments on six open-weight LLMs show safety behaviors evolve gradually across threat levels rather than flipping from refusal to compliance. Models with similar Attack Success Rates can exhibit distinct response distributions, arguing for behavior-aware safety evaluation.
IdeaAMBIG benchmark with 660 instances measures whether LLMs can spot and fix underspecified research-method details for faithful implementation.
Researchers introduce IdeaAMBIG, a benchmark of 660 evidence-grounded instances (163 real-world gaps from reproducibility reports and GitHub issues, 497 controlled synthetic gaps) built from papers, codebases, and reproduction artifacts. It evaluates codification-readiness assessment, defect localization, and clarification action generation. Across 13 LLMs, the best model achieved only 9.6% Macro Defect Recovery Rate on real-world instances but 80.6% clarification success when given the annotated defect. An oracle study showed gold resolutions raise the codification-ready rate from 14% to 98%, identifying defect localization as the main bottleneck.
IdeaAMBIG benchmark of 660 specification-gap instances shows LLMs localize implementation-critical research gaps poorly, with best model at 9.6% defect recovery.
IdeaAMBIG is a benchmark of 660 evidence-grounded instances evaluating whether research-method specifications provide enough information for faithful implementation: 163 real-world gaps from reproducibility reports and GitHub issues plus 497 controlled synthetic gaps. It tests codification-readiness assessment, defect localization, and clarification action generation across 13 LLMs. The best model achieves only a 9.6% Macro Defect Recovery Rate on real-world instances, though 80.6% clarification success when given the annotated defect, and an oracle study shows gold resolutions raise codification-ready rates from 14% to 98%. Defect localization emerges as the main bottleneck across all evaluated models.
xDailyBench tests 11 frontier LLMs on 248 real-life consultation tasks; the best models score 75.6% and lag on implicit requirements.
The benchmark spans 51 scenarios across personal life, white-collar work, learning and research, and cross-domain activities, grounded in requests users actually completed or intended to complete with AI. Tasks are scored with fine-grained binary rubrics covering explicit and implicit requirements under standardized agentic settings. Across 11 frontier models, the best achieved a 75.6% task-level score, with all models performing at least 9 percentage points worse on implicit than explicit requirements.
Introduces VEX-Bench, 75 expert-labeled real-world cases testing whether LLM agents can assess supply chain vulnerability exploitability; frontier models reach about 80% F1.
VEX-Bench is the first benchmark evaluating LLM agents on assessing whether upstream dependency vulnerabilities are exploitable in downstream projects, with 75 real-world expert-labeled cases across Python, Java, and Go mined from GitHub. Nine models across three agent harnesses were evaluated; GPT-5.5 and Claude Opus 4.6 reach approximately 80% F1 on binary vulnerability-status classification, but only GPT-5.5 surpasses 70% macro-F1 on fine-grained justification classification. The gap highlights the difficulty of moving beyond binary exploitability calls to explaining exploitability reasons, unlike prior benchmarks targeting zero-day settings.
Simon Willison released llm-openrouter 0.7.1, a performance fix for loading OpenRouter models in his LLM CLI plugin.
Version 0.7.1 of the llm-openrouter plugin addresses a performance problem when loading OpenRouter models in the LLM command-line tool. The fix was contributed by GitHub user waveplate. It is a minor maintenance release with no security implications noted.
Independent developer Hugo Vergnes trained a 3.8B-parameter Llama-style model to 0.384 CORE on 65B tokens for $998 in 43 hours on rented B200s.
Hugo Vergnes trained little-lm, a 3.848B-parameter decoder-only LLM, on 65.3B tokens in 43 hours for $998 using rented NVIDIA B200s, scoring 0.384 on the CORE benchmark and beating nanochat d32 (0.310) at similar cost. The Llama-style architecture uses RMSNorm, RoPE, GQA with 24 query and 8 KV heads, relu-squared MLPs, QK-norm, and ResFormer-style value embeddings that account for 19% of parameters. Key wins included the Muon optimizer for matrix parameters, a trapezoidal learning-rate schedule with linear cooldown, FP8 training plus vocabulary padding for roughly 33% throughput gains, and the ClimMix dataset over FineWeb-Edu. The project, inspired by Karpathy's nanochat, was built as a config-driven YAML framework for small LLM training.
Benchmark Radar provides a living searchable database of 1,283 AI benchmark records and 12,916 score observations drawn from 37 daily discovery sources.
Benchmark Radar combines daily discovery of benchmark papers, repositories, datasets, and releases from 13 direct connectors and 24 first-party feeds into a searchable catalog with model card mentions and score histories. The catalog contains 1,283 source records drawn from 4 benchmark catalogs plus 12,916 numeric observations on 790 records. The release includes a web dashboard with leaderboard, Pareto frontier of score versus usage, saturation and trend views, daily feeds, a CLI, and reproducible analysis. The paper audits the full catalog and examines benchmark saturation and limits of score comparisons.
A methodological study shows close LLM-agent leaderboard rank gaps on SWE-bench and similar benchmarks often do not support superiority claims.
The paper defines an estimand-aware pairwise procedure for comparing agents, checking common support and applying explicit uncertainty rules and practical margins. Across SWE-bench, AgentRewardBench, and tau2-bench, close rank differences are frequently unresolved, and proxy labels or utility rules can change which system is selected. The authors argue a leaderboard score summarizes a released evaluation but does not by itself justify pairwise superiority conclusions.
Researchers propose auditing LLM bias via relative hidden-state representations, detecting bias increases with 3-50x less compute than output-level benchmarks.
The paper introduces a reference-based bias auditing method that compares hidden-state representations across model variants, such as before and after fine-tuning, by encoding sentences relative to a fixed anchor set. The resulting Representational Bias Shift (Delta-B) correlates with output-level bias change in 15 of 18 tested settings, reaching |r| = 0.84 under full fine-tuning across WildGuardMix, DecodingTrust, and ToxiGen benchmarks. Thresholding Delta-B detects checkpoints whose bias increased with ROC AUC between 0.65 and 0.99 and beats a SEAT-based baseline, while auditing a model in about three minutes with 3-50x less compute.
VDiff-Bench, a 1,756-question benchmark, shows multimodal LLMs struggle with fine-grained image-difference identification, scoring as low as 8.7% on low-level changes.
VDiff-Bench is a multiple-choice benchmark of 1,756 four-way questions over image pairs covering 10 change categories including position, motion, color, texture, OCR/text and illumination, with curated hard negatives. Evaluation of 11 state-of-the-art open- and closed-source MLLMs shows fine-grained visual comparison remains brittle: 7-8B-scale open-source models score 52.5-70.6% on semantic changes but only 8.7-33.3% on low-level changes like noise and texture. Notably, Grok 4.3 shows a sharp performance drop on noise and texture differences, falling behind large open-source models like Kimi K2.5 and K3.
Bruce Schneier highlights two papers: the CIMemories benchmark shows frontier LLMs leak memory attributes up to 69%, and an RL method reduces inappropriate disclosures.
Bruce Schneier discusses contextual integrity in LLMs, referencing the CIMemories benchmark, which uses synthetic profiles with 100+ attributes per user to test whether models with persistent memory disclose sensitive information appropriately. Evaluation showed frontier models exhibit up to 69% attribute-level violations, with GPT-5's violation rate rising from 0.1% to 9.6% across 40 tasks and reaching 25.1% with repeated prompting, showing unstable leakage behavior. A second paper introduces a reinforcement learning framework trained on a synthetic 700-example dataset that substantially reduces inappropriate disclosure while maintaining task performance, with improvements transferring to the human-annotated PrivacyLens benchmark.
Researchers release Phi-Bench, a benchmark evaluating frontier LLMs on open-ended, long-horizon engineering and optimization of the LLM infrastructure stack.
Phi-Bench evaluates LLMs on open-ended engineering of the LLM infrastructure stack, derived from optimization problems studied in frontier research and grounded in real-world code repositories. Tasks range from localized kernel-level function completion to long-horizon implementation and end-to-end system optimization. Experiments on frontier LLMs reveal current capabilities and limitations on the path toward autonomous optimization of future AI infrastructure.
A new benchmark shows LLMs reach 68.3-93% accuracy propagating local revisions across conversationally generated artifacts, with parallel-sample selection most cost-effective.
The paper introduces a benchmark for revision propagation: when users request a local change, LLMs must identify dependencies and update all affected parts of an artifact generated through conversation, where context lives in the chat history. Nine revision methods, including sequential reflection and parallel sampling variants, were evaluated on gpt-oss-20b/120b, gpt-5.4-mini, and qwen3.5-9b/27b/122b. Baselines scored 68.3-93% accuracy, and selecting among three parallel samples via LLM-based or medoid selection improved accuracy by 2.2-9.7% as the most cost-effective test-time compute strategy. Code and dataset are released.
Uno pairs autoregressive LLMs with lightweight diffusion weights to draw multiple tokens in parallel, delivering up to 3x lossless speedup without a draft model.
The paper introduces diffusion-augmented LLMs: autoregressive weights trained with the standard next-token objective plus lightweight diffusion weights trained via a Diffusion Distillation phase to emit multiple tokens in parallel. Psi-Spec samplers enable lossless acceleration without the separate draft model required by speculative decoding. The 8B Uno model outperforms the 26B open DiffusionGemma and proprietary Mercury 2 on agentic tool use, coding, and long-context reasoning benchmarks, with up to 3x throughput gains over the base model at all evaluated batch sizes. Code and checkpoints are released publicly.
TempCloze benchmark tests Video-LLMs' temporal reasoning with 1,521 videos, finding temporal alignment is the primary failure mode across 31 models.
TempCloze is a video cloze benchmark in which models must identify the true missing middle clip given the beginning and ending clips, using 1,521 carefully filtered videos from seven sources, mostly long-take and egocentric footage. Distractors are constructed along three dimensions: Semantic, Alignment and Progression, with shared scenes and objects to reduce appearance cues. Evaluation of 10 proprietary and 21 open-source Video-LLMs found Alignment is the primary bottleneck, with models often recognizing plausible semantics and local event progression but struggling with temporal alignment.
ByteDance Seed's HarnessDev benchmark finds LLM-built agent harnesses trail human engineering on code and search, with only 34 of 64 revisions generalizing.
Researchers from ByteDance Seed, SUTD, Georgia Tech, M-A-P, and TokenWave.AI introduce HarnessDev, a benchmark that evaluates the runnable agent harness an LLM writes rather than its answers, using Creation and Evolution stages across SWE-bench Pro, Terminal-Bench 2.1, MLE-bench, EQ-Bench3, and BrowseComp (2,207 instances). Six creator models including Opus 4.8, GPT-5.5, Gemini 3.1 Pro, DeepSeek V4 Pro, Qwen 3.7 Max, and Seed 2.0 Pro were tested; Opus 4.8 posted the best average of 67.8 versus an 86.2 human-engineered reference. Self-built harnesses beat references on writing and ML experimentation but lag badly on code and search, and quality proved executor-specific: Opus 4.8 fell from 69.3 to 33.0 on SWE-bench Pro when the executor was switched to Gemini. Evolution gains were small and noisy: of 64 adjacent changes, feedback and held-out scores agreed only 34 times (53.1%), and much generated state and memory code never executed.
Elo-per-token analysis shows LLM agents' marginal gains drop below independent sampling at scale; parallel sessions beat one long session.
The paper proposes Elo-per-token analysis, using a Bradley-Terry model to measure how agent performance scales with token budget on open-ended tasks with continuous scoring. Across four agents and four benchmarks with sessions up to 100M tokens, agents initially convert tokens to Elo faster than independent sampling but eventually slow below the linear-in-log-compute reference. The authors define a scaling inflection point and show that splitting 100M tokens across parallel sessions on FrontierCS Polyomino Packing gains +264 Elo over one long session and +355 over ten short sessions. Human contestants on shared AtCoder Heuristic Contest tasks improve superlinearly, indicating headroom over current agents.
WearableQA benchmark tests LLM health reasoning over longitudinal wearable data; the best of 14 evaluated LLMs reaches 72.9% accuracy.
WearableQA comprises 4,084 ten-option multiple-choice questions built from wearable time series, blood biomarkers, and demographics of 200 real users with up to 500 days of daily measurements. It defines 16 question types along two axes: data versus health reasoning, and single- versus cross-signal reasoning. Evaluation of 14 proprietary and open-source LLMs shows performance from 19.6% to 72.9% against a 10% chance baseline, with most models below 60%.
Gavel reads native skill-routing signals from a frozen LLM's forward passes with two linear maps, beating retrieve-and-rerank pipelines by up to 21.9 points on Qwen3-32B.
Gavel (Glance And Verdict from a frozen LLM) elicits skill routing from a frozen agent LLM using two trained linear maps that read mid-layer states, keeping all skill text out of context. A glance step scores the full library against compact per-skill banks built in one forward pass at installation; a verdict step resumes shortlisted skills' forward passes and fuses likelihood and yes/no judgments as a product of experts. It transfers zero-shot to three public benchmarks plus SkillTraj, a new benchmark of 372 simulated agent trajectories. On Qwen3-32B it beats progressive disclosure and retrieve-and-rerank pipelines adding 1.2B–16B external parameters by up to 13.4 points on written tasks and 21.9 when skills are needed mid-rollout.
A study finds LLM-synthesized CodeQL queries improve average F1-score by 82% over baseline queries, offering scalable vulnerability detection versus direct LLM scanning.
Researchers conducted an empirical study evaluating whether LLMs can synthesize executable CodeQL queries from National Vulnerability Database vulnerability data. LLM-generated queries significantly enhanced baseline CodeQL suites, yielding an 82% improvement in average F1-score across a diverse set of real-world vulnerabilities. A cost-benefit analysis shows direct LLM-based scanning of entire repositories is often computationally and financially prohibitive, while LLM query synthesis offers a scalable and cost-effective alternative for large-scale vulnerability detection.
Researchers introduce GRAF, a greedy framework for crowdsourcing contest self-selection, and LLMScore, an LLM-driven method that auto-designs its scoring algorithm.
The paper studies self-selection in Tullock contests (SSTC), where workers choose contests and then compete within them. GRAF is a greedy polynomial-time framework that orders workers by a score vector with zero worker regret and platform optimality guarantees in special cases. LLMScore is an LLM-driven evolutionary framework that produces human-readable, inspectable scoring code, jointly optimizing platform utility and worker satisfaction. Across 1,000 synthetic instances in four settings, GRAF with LLMScore achieves high-quality, often near-optimal outcomes with low worker regret, transferring from small training instances to larger, structurally different settings.
Systematic review of 66 studies finds LLMs for HVAC operations are mostly research-stage, with no ready-now deployment and only four pilot-level studies.
A critical review of 66 peer-reviewed studies from 2023 to March 2026 examines LLMs for HVAC operations in building energy systems. Only four studies reach pilot-level evidence, none reports sustained operational deployment, and 63 of 66 are research-only. Conventional ML, MPC, and RL remain dominant for high-frequency control and short-horizon forecasting, and the evidence supports LLMs primarily as semantic and workflow layers rather than autonomous controllers.
Researchers release PIA-Bench, the first open benchmark evaluating how accurately LLMs can automate privacy impact assessments using 73 curated federal PIAs.
PIA-Bench is the first open benchmark for evaluating large language models on real-world privacy impact assessments (PIAs). The authors audited 499 expert-authored PIAs published by US federal agencies and curated 73 structured PIAs comprising 451 privacy risk items and 831 mitigation items. Off-the-shelf LLMs were found to produce meaningful assessments while identifying clear avenues for improvement. The paper calls for domain-specific LLM agent workflows, accountable LLM infrastructure, and new quality standards for PIAs.
Amazon ICML paper uses Ising models to correct correlated LLM-judge votes, beating accuracy-weighted panels by 9-14%.
Amazon Science describes an ICML paper, "Dependence-aware label aggregation for LLM-as-a-judge via Ising models," addressing how correlated judge outputs inflate majority-vote confidence. The unsupervised method models pairwise dependence between judges, learning both reliability and similarity without human reference labels. Tested on relevance, toxicity, and summarization tasks with 10 judge models at temperature zero, it outperformed accuracy-weighted voting by 9% to 14%.
Hacker News · AI