Negative narrative immersion causes 12-31% drops in LLM moral accuracy and produces structured shifts that appear in downstream applications.
super hub Mixed citations
Measuring Massive Multitask Language Understanding
Mixed citation behavior. Most common role is background (45%).
445
Pith papers citing it
Background
45% of classified citations
abstract
We propose a new test to measure a text model's multitask accuracy. The test covers 57 tasks including elementary mathematics, US history, computer science, law, and more. To attain high accuracy on this test, models must possess extensive world knowledge and problem solving ability. We find that while most recent models have near random-chance accuracy, the very largest GPT-3 model improves over random chance by almost 20 percentage points on average. However, on every one of the 57 tasks, the best models still need substantial improvements before they can reach expert-level accuracy. Models also have lopsided performance and frequently do not know when they are wrong. Worse, they still have near-random accuracy on some socially important subjects such as morality and law. By comprehensively evaluating the breadth and depth of a model's academic and professional understanding, our test can be used to analyze models across many tasks and to identify important shortcomings.
hub tools
citation-role summary
background 30
dataset 29
method 5
baseline 3
citation-polarity summary
claims ledger
- abstract We propose a new test to measure a text model's multitask accuracy. The test covers 57 tasks including elementary mathematics, US history, computer science, law, and more. To attain high accuracy on this test, models must possess extensive world knowledge and problem solving ability. We find that while most recent models have near random-chance accuracy, the very largest GPT-3 model improves over random chance by almost 20 percentage points on average. However, on every one of the 57 tasks, the best models still need substantial improvements before they can reach expert-level accuracy. Models
authors
co-cited works
representative citing papers
Presents the first fully open pipeline for clinical LLMs by unifying eight public QA datasets with three clinician-vetted synthetic extensions and applying it to five base models to achieve benchmark gains while maintaining auditability.
AI model builders mostly highlight unique benchmarks that act as flexible narrative tools for market positioning rather than standardized scientific measurements.
HodgeCover isolates the harmonic kernel of a simplicial Laplacian on an expert 2-complex to identify irreducible merge cycles and selects experts for aggressive compression, matching or exceeding baselines on open-weight MoE models.
EnergyAgentBench is a new benchmark with 70 task variants that evaluates LLM agents on live energy data for datacenter siting, long-horizon optimization, and causal grid diagnosis.
LCDD creates sparse carriers for SFT behaviors that SFT-Eraser can reverse, with ablations showing the sparse structure enables causal control.
ArgBench unifies 33 existing datasets into a standardized benchmark for testing LLMs across 46 argumentation tasks and analyzes the impact of prompting techniques and model factors on performance.
MCP-Atlas is a new benchmark with 1000 tasks on production MCP servers that uses claim-level scoring to evaluate LLM agents on realistic multi-step tool-use competency.
LLaDA is a scalable diffusion-based language model that matches autoregressive LLMs like LLaMA3 8B on tasks and surpasses GPT-4o on reversal poem completion.
AgentClinic is a multimodal agent benchmark demonstrating that LLM diagnostic accuracy on MedQA drops to below one-tenth in sequential clinical simulations, with Claude-3.5 leading and large tool-use differences across models.
FlipGuard perturbs LLM weights prior to quantization to neutralize quantization-conditioned backdoor attacks, evaluated via the Defense Effectiveness Ratio on multiple models and quantization schemes.
LLM agents often fail to abstain at the right time in uncertain multi-turn tasks, and the CONVOLVE context engineering method raises timely abstention rates on WebShop from 26.7 to 57.4 without parameter updates.
Face-Feature Tuning is a label-free logit remapping method that reduces FPR/TPR gaps across groups in deepfake detection while preserving overall accuracy.
A new fault-injection framework enables a systematic empirical study that produces 17 takeaways on error propagation in LLM inference and four software-only mitigation directions.
CultureForest benchmark shows top LLMs degrade sharply on open-ended cultural reasoning tasks, exhibit regional disparities, and are limited more by effective use of knowledge than by lack of knowledge itself.
THRD introduces a training-free multi-turn defense framework that models temporal risk accumulation to reduce jailbreak attack success rates to 0.2-4.0% on LLMs with under 1.5% utility degradation.
A bipartite factor graph with message-passing protocol and asymmetric damping aggregates multi-LLM predictions, cutting token use by 97% and API calls by 6X while outperforming baselines on MMLU, MMLU-Pro, GPQA, and MedMCQA.
RHELM is a benchmark for LLM long-term memory with dynamic profiles, heterogeneous sources, and 27 memory characteristics that reveals weaknesses in existing models for multi-source aggregation and contextual reasoning.
ReactBench is a new benchmark with four cause-targeted tasks that uses adversarial images, hallucination-inducing queries, and Chain-of-Thought analysis to expose specific failure modes in current multimodal large language models.
K-FinHallu is the first multi-turn Korean financial RAG hallucination benchmark; frontier LLMs struggle especially on justified abstention while an 8B fine-tuned model reaches competitive performance.
ConMoE consolidates MoE experts into a smaller prototype pool via deterministic remapping based on contribution and replaceability, matching or beating pruning/merging baselines at 25-50% reduction on three models.
LLMs struggle to associate epistemic markers with stable internal confidence levels across distributions, even under model-centric interpretations, while maintaining somewhat consistent marker rankings.
SiDP distributes model weights across a DP group with WaS and CaS modes to increase KV cache capacity by up to 1.8x and end-to-end throughput by up to 1.5x over vLLM on H20/H200/B200 GPUs for offline LLM inference.
Introduces a three-tier architecture with an agent runtime layer and four primitives for agent-aware policies in LLM serving, validated on KV caching via CacheSage showing 13-37pp hit-rate gains on five workloads.
citing papers explorer
-
MoE-Prefill: Zero Redundancy Overheads in MoE Prefill Serving
MoE-Prefill achieves 1.35-1.59x higher throughput for prefill-only MoE serving by using asynchronous expert parallelism to overlap weight AllGather with computation and prefix-aware routing with true-FLOPs tracking.
-
SciEval: A Benchmark for Automatic Evaluation of K-12 Science Instructional Materials
SciEval is a new benchmark of expert-annotated K-12 science lessons for LLM-based automatic evaluation, where zero-shot models perform poorly but fine-tuning yields up to 11% gains.
-
Analysis and Explainability of LLMs Via Evolutionary Methods
Evolutionary trees from LLM weights recover ground-truth training topologies and identify key datasets and layers through phenotypic analysis.
-
Green Shielding: A User-Centric Approach Towards Trustworthy AI
Green Shielding introduces CUE criteria and the HCM-Dx benchmark to demonstrate that routine prompt variations systematically alter LLM diagnostic behavior along clinically relevant dimensions, producing Pareto-like tradeoffs in plausibility versus coverage.
-
Breaking the Secret: Economic Interventions for Combating Collusion in Embodied Multi-Agent Systems
A mutagenic incentive mechanism reshapes payoffs in embodied MAS to induce strategic defection from collusion, achieving performance comparable to non-collusion baselines in simulations and real-world tests.
-
Are LLM Uncertainty and Correctness Encoded by the Same Features? A Functional Dissociation via Sparse Autoencoders
Uncertainty and correctness in LLMs are encoded by distinct feature populations, with suppression of confounded features improving accuracy and reducing entropy.
-
How Tokenization Limits Phonological Knowledge Representation in Language Models and How to Improve Them
Subword tokenization impairs phonological knowledge encoding in LMs, but an IPA-based fine-tuning method restores it with minimal impact on other capabilities.
-
Pushing the Boundaries of Multiple Choice Evaluation to One Hundred Options
Scaling multiple-choice questions to 100 options on a Korean error detection task shows that LLM performance on conventional benchmarks overstates true competence due to shortcut strategies.
-
AgileLog: A Forkable Shared Log for Agents on Data Streams
AgileLog introduces forkable shared logs with cheap forking and isolation to support AI agents on data streams.
-
Awakening Dormant Experts:Counterfactual Routing to Mitigate MoE Hallucinations
Counterfactual Routing awakens dormant experts in MoE models via layer-wise perturbation and a new CEI metric, raising factual accuracy 3.1% on average across TruthfulQA, FACTOR, and TriviaQA without extra inference cost.
-
Retrieval as Generation: A Unified Framework with Self-Triggered Information Planning
GRIP integrates retrieval into autoregressive generation through self-triggered control tokens for dynamic query planning, outperforming RAG baselines on QA benchmarks with fewer parameters than GPT-4o.
-
EgoTL: Egocentric Think-Aloud Chains for Long-Horizon Tasks
EgoTL provides a new egocentric dataset with think-aloud chains and metric labels that benchmarks VLMs on long-horizon tasks and improves their planning, reasoning, and spatial grounding after finetuning.
-
GeoMMBench and GeoMMAgent: Toward Expert-Level Multimodal Intelligence in Geoscience and Remote Sensing
GeoMMBench reveals deficiencies in current multimodal LLMs for geoscience tasks while GeoMMAgent demonstrates that tool-integrated agents achieve significantly higher performance.
-
Every Response Counts: Quantifying Uncertainty of LLM-based Multi-Agent Systems through Tensor Decomposition
MATU quantifies uncertainty in LLM multi-agent systems by turning reasoning trajectories into embedding matrices, stacking runs into a tensor, and decomposing it to separate sources of variability.
-
Social Dynamics as Critical Vulnerabilities that Undermine Objective Decision-Making in LLM Collectives
Social dynamics in LLM collectives cause representative agents to make less accurate decisions as peer pressure increases through larger adversarial groups, more capable peers, longer arguments, and persuasive styles.
-
FrontierFinance: A Long-Horizon Computer-Use Benchmark of Real-World Financial Tasks
FrontierFinance benchmark shows human financial experts outperform state-of-the-art LLMs by achieving higher scores and more client-ready outputs on realistic long-horizon tasks.
-
DeonticBench: A Benchmark for Reasoning over Rules
DEONTICBENCH is a new benchmark of 6,232 deontic reasoning tasks from U.S. legal domains where frontier LLMs reach only ~45% accuracy and symbolic Prolog assistance plus RL training still fail to solve tasks reliably.
-
PolyReal: A Benchmark for Real-World Polymer Science Workflows
PolyReal benchmark shows leading MLLMs perform well on polymer knowledge reasoning but drop sharply on practical tasks like lab safety analysis and raw data extraction.
-
Bridging the Missing-Modality Gap: Improving Text-Only Calibration of Vision Language Models
A new Latent Imagination Module uses cross-attention to predict latent visual embeddings from text, improving accuracy and calibration of vision-language models on text-only inputs.
-
A Switch-Centric In-Network Architecture for Accelerating LLM Inference in Shared-Memory Network
SCIN uses an in-switch accelerator for direct memory access and 8-bit in-network quantization during All-Reduce, delivering up to 8.7x faster small-message reduction and 1.74x TTFT speedup on LLaMA-2 models.
-
Path-Constrained Mixture-of-Experts
PathMoE constrains expert paths in MoE models by sharing router parameters across layer blocks, yielding more concentrated paths, better performance on perplexity and tasks, and no need for auxiliary losses.
-
KMMMU: Evaluation of Massive Multi-discipline Multimodal Understanding in Korean Language and Context
KMMMU benchmark demonstrates that leading multimodal models achieve at most 52.42% accuracy on hard Korean exam questions, highlighting limitations in non-English multimodal understanding.
-
PEEM: Prompt Engineering Evaluation Metrics for Interpretable Joint Evaluation of Prompts and Responses
PEEM is a multi-criteria LLM-based evaluator for prompts and responses that aligns with standard accuracy while enabling zero-shot prompt optimization via feedback.
-
EvoESAP: Non-Uniform Expert Pruning for Sparse MoE
EvoESAP uses evolutionary search guided by a speculative-decoding-inspired ESAP metric to discover non-uniform layer-wise sparsity allocations for MoE expert pruning, improving generation accuracy up to 19.6% at 50% sparsity.
-
VISTA-Bench: Do Vision-Language Models Really Understand Visualized Text as Well as Pure Text?
VISTA-Bench shows vision-language models degrade on visualized text in images compared to equivalent pure text, with larger gaps under increased perceptual difficulty.
-
Towards Distillation-Resistant Large Language Models: An Information-Theoretic Perspective
A learned transformation matrix minimizes CMI in teacher logits to degrade distillation performance while preserving task accuracy.
-
Norm Anchors Make Model Edits Last
Norm-Anchor Scaling breaks the norm-feedback loop in sequential LLM editing by anchoring value vectors to original norms, improving long-run performance by 72.2% and extending the editing horizon over 4x.
-
Results-Actionability Gap: Understanding How Practitioners Evaluate LLM Products in the Wild
Qualitative study of 19 practitioners reveals ten LLM product evaluation practices and introduces the results-actionability gap as a key barrier to turning findings into improvements.
-
Batch-of-Thought: Cross-Instance Learning for Enhanced LLM Reasoning
Batch-of-Thought enables cross-instance learning by jointly processing related queries in batches, yielding higher accuracy and up to 61% lower inference costs on LLM reasoning tasks.
-
CORE: Concept-Oriented Reinforcement for Bridging the Definition-Application Gap in Mathematical Reasoning
CORE is a concept-oriented RL method that synthesizes quizzes, injects concept snippets into rollouts, and reinforces conceptual trajectories to close the gap between restating definitions and applying them in math problems.
-
MIDUS: Memory-Infused Depth Up-Scaling
MIDUS replaces duplicated FFN branches in depth up-scaling with head-wise memory layers using product-key retrieval and HIVE to deliver lightweight, head-conditioned residual capacity.
-
Data Mixing Agent: Learning to Re-weight Domains for Continual Pre-training
An RL agent learns domain re-weighting policies from evaluation feedback to improve balanced performance in continual pre-training of LLMs across source and target domains.
-
PRIMETIME : Limits of LLMs in Temporal Primitives
PRIMETIME generator reveals that LLM datetime parsing and arithmetic primitives are individually unreliable but fully learnable via fine-tuning, enabling frontier-level accuracy on event planning with small LoRA models.
-
L1: Controlling How Long A Reasoning Model Thinks With Reinforcement Learning
LCPO trains L1 reasoning models to adhere to prompt-specified CoT lengths, supporting accuracy-compute trade-offs and yielding short reasoning models that outperform larger baselines at matched lengths.
-
Stay Focused: Problem Drift in Multi-Agent Debate
The paper defines and measures 'problem drift' in multi-agent LLM debates across tasks and proposes DRIFTJudge and DRIFTPolicy as baselines to detect and reduce it.
-
Semantic Integrity Matters: Benchmarking and Preserving High-Density Reasoning in KV Cache Compression
KV cache compression causes task-dependent degradation in high-density reasoning due to disrupted CoT links; ShotKV mitigates this by preserving few-shot examples as indivisible semantic units through phase separation, delivering 9-18% accuracy gains and 11% latency reduction.
-
FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI
FrontierMath is a new benchmark of hundreds of original hard math problems that current AI models solve less than 2% of.
-
RiTeK: A Dataset for Large Language Models Complex Reasoning over Textual Knowledge Graphs in Medicine
Introduces RiTeK dataset for complex LLM reasoning over medical TKGs with expert-validated queries and shows existing retrievers struggle on the benchmark.
-
Functional-level Uncertainty Quantification for Calibrated Fine-tuning on LLMs
UQ4CT integrates functional-level uncertainty calibration into mixture-of-experts LoRA fine-tuning via a dedicated loss, cutting expected calibration error by over 25% on multiple-choice and generative QA tasks.
-
Moshi: a speech-text foundation model for real-time dialogue
Moshi is the first real-time full-duplex spoken large language model that casts dialogue as speech-to-speech generation using parallel audio streams and an inner monologue of time-aligned text tokens.
-
V-RoAst: Visual Road Assessment. Can VLM be a Road Safety Assessor Using the iRAP Standard?
V-RoAst applies zero-shot VLMs (Gemini-1.5-flash, GPT-4o-mini) to iRAP road safety attribute classification on a new ThaiRAP image dataset and compares them to CNN baselines, finding better generalization to unseen classes but weaker spatial reasoning.
-
Refusal in Language Models Is Mediated by a Single Direction
Refusal in language models is mediated by a single direction in residual stream activations that can be erased to disable safety or added to elicit refusal.
-
Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing
Magpie synthesizes 300K high-quality alignment instructions from Llama-3-Instruct via auto-regressive prompting on partial templates, enabling fine-tuned models to match official instruct performance on AlpacaEval, ArenaHard, and WildBench.
-
DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model
DeepSeek-V2 delivers top-tier open-source LLM performance using only 21B active parameters by compressing the KV cache 93.3% and cutting training costs 42.5% via MLA and DeepSeekMoE.
-
LongRoPE: Extending LLM Context Window Beyond 2 Million Tokens
LongRoPE extends LLM context windows to 2048k tokens via search for non-uniform positional interpolation, progressive fine-tuning from 256k, and short-context readjustment.
-
Self-Rewarding Language Models
Iterative self-rewarding via LLM-as-Judge in DPO training on Llama 2 70B improves instruction following and self-evaluation, outperforming GPT-4 on AlpacaEval 2.0.
-
A Rank Stabilization Scaling Factor for Fine-Tuning with LoRA
LoRA adapters should be scaled by 1/sqrt(rank) rather than 1/rank to stabilize learning and enable effective use of higher ranks during fine-tuning of large language models.
-
WizardLM: Empowering large pre-trained language models to follow complex instructions
WizardLM uses LLM-driven iterative rewriting to generate complex instruction data and fine-tunes LLaMA to reach over 90% of ChatGPT capacity on 17 of 29 evaluated skills.
-
Capabilities of GPT-4 on Medical Challenge Problems
GPT-4 exceeds the USMLE passing score by more than 20 points and outperforms both GPT-3.5 and the medically fine-tuned Med-PaLM on the MultiMedQA benchmarks.
-
Reinforcement Learning with Metacognitive Feedback Elicits Faithful Uncertainty Expression in LLMs
RLMF uses quality of model self-judgments to refine RL rankings and select training data, achieving SOTA faithful calibration while preserving accuracy and outperforming standard RL by up to 63%.