Introduces the MMLU benchmark of 57 tasks and shows that current models, including GPT-3, achieve low accuracy far below expert level across academic and professional domains.
hub Mixed citations
PIQA: Reasoning about Physical Commonsense in Natural Language
Mixed citation behavior. Most common role is background (56%).
41
Pith papers citing it
Background
56% of classified citations
abstract
To apply eyeshadow without a brush, should I use a cotton swab or a toothpick? Questions requiring this kind of physical commonsense pose a challenge to today's natural language understanding systems. While recent pretrained models (such as BERT) have made progress on question answering over more abstract domains - such as news articles and encyclopedia entries, where text is plentiful - in more physical domains, text is inherently limited due to reporting bias. Can AI systems learn to reliably answer physical common-sense questions without experiencing the physical world? In this paper, we introduce the task of physical commonsense reasoning and a corresponding benchmark dataset Physical Interaction: Question Answering or PIQA. Though humans find the dataset easy (95% accuracy), large pretrained models struggle (77%). We provide analysis about the dimensions of knowledge that existing models lack, which offers significant opportunities for future research.
hub tools
citation-role summary
background 5
dataset 3
baseline 1
citation-polarity summary
representative citing papers
GPT-3 shows that scaling an autoregressive language model to 175 billion parameters enables strong few-shot performance across diverse NLP tasks via in-context prompting without fine-tuning.
HERMES provides a reusable hierarchical labeling substrate for pre-training data that reveals granularity-specific effects in data mixing rules during model training.
2-bit quantized reasoning models exhibit process failures like loops and delayed commitment that degrade end-to-end performance, but FP16 planning and loop rescue recover accuracy on MATH-500 from 17.2% to 74.2% for Qwen3-8B while retaining speed gains.
LoopUS converts pretrained LLMs into looped latent refinement models via block decomposition, selective gating, random deep supervision, and confidence-based early exiting to improve reasoning performance.
Token-to-Mask remasking improves self-correction in diffusion LLMs by resetting erroneous commitments to masks rather than overwriting them, yielding +13.33 points on AIME 2025 and +8.56 on CMATH.
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.
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.
BitNet b1.58 shows that ternary 1.58-bit LLMs can match full-precision performance at substantially lower inference cost.
Massive activations are constant large values in LLMs that function as indispensable bias terms and concentrate attention probabilities on specific tokens.
SCAPE enables 90-99% sparse gradient communication in sharded Adam-style LLM training by deriving masks from first-moment statistics, achieving up to 43.3% faster pre-training on Llama-500M with no loss in validation loss or downstream accuracy.
One-step gradient delay is optimizer-dependent rather than intrinsically unstable, with Muon and error-feedback correction enabling async pipeline parallelism to match synchronous performance on models up to 10B parameters.
R2LM combines causal attention with a reverse Mamba SSM sidecar to supply right-side context in dLLMs, claiming 2.4x-12.9x throughput gains over bidirectional dLLMs and 1.9x-2.9x over AR baselines while matching or exceeding quality.
Derives an upper bound on frozen LM expected risk from proxy risk, SAE reconstruction gap, concept-pool mismatch and sparse complexity, with non-vacuous bounds observed on GPT-2, Gemma-2B and Llama-3-8B.
Self-generated T2T training on LLaDA2.1-mini improves benchmark accuracy and lowers edit intensity by supervising recovery from model-generated corruptions instead of random ones.
TWLA is a PTQ method using E2M-ATQ, KOTMS, and ILA-AMP to enable W1.58A4 quantization for LLMs with maintained accuracy.
STAR rethinks MoE routing as structure-aware subspace learning by adding a GHA-tracked principal subspace to standard routers, yielding more stable specialization and better performance on synthetic, language, and vision tasks.
Deeper transformer layers benefit from context-free token-specific value vectors in a Bank of Values lookup table, improving performance over standard attention with less compute.
LINK improves cross-lingual knowledge transfer via lexical substitutions in English pretraining data, yielding notable downstream gains and up to 2x training speedup across eight languages and five model sizes.
In-Place TTT adapts LLM MLP projection matrices at test time with a next-token-aligned objective and chunk-wise updates, enabling better long-context performance as a drop-in enhancement.
Short sliding windows in hybrid attention-xLSTM models boost long-context performance by encouraging long-term memory use, and stochastic window sizing improves both short and long tasks.
HyperAdapt performs parameter-efficient fine-tuning by row- and column-wise diagonal scaling to induce high-rank updates with only n+m trainable parameters.
MoE models with activation rates in an optimal region outperform dense LLMs of identical total parameter count, training compute, and data budget, with the optimal region consistent across scales.
Pretraining data determines loss-to-loss scaling laws in LLMs, while model size, optimization, tokenizer, and architecture have limited impact.
citing papers explorer
-
Measuring Massive Multitask Language Understanding
Introduces the MMLU benchmark of 57 tasks and shows that current models, including GPT-3, achieve low accuracy far below expert level across academic and professional domains.
-
Language Models are Few-Shot Learners
GPT-3 shows that scaling an autoregressive language model to 175 billion parameters enables strong few-shot performance across diverse NLP tasks via in-context prompting without fine-tuning.
-
Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone
Phi-3-mini (3.8B params, 3.3T tokens) reaches 69% MMLU and 8.38 MT-bench, matching larger models, with scaled-up 7B/14B variants and phi-3.5 extensions for multilingual, MoE, and vision capabilities.
-
PaLM: Scaling Language Modeling with Pathways
PaLM 540B demonstrates continued scaling benefits by setting new few-shot SOTA results on hundreds of benchmarks and outperforming humans on BIG-bench.
-
Gemma 3 Technical Report
Gemma 3 introduces multimodal open models with architectural changes for efficient long context, trained via distillation and a new post-training recipe that makes the 4B version competitive with prior 27B models and the 27B version comparable to Gemini-1.5-Pro.
-
Gemma: Open Models Based on Gemini Research and Technology
Gemma introduces open 2B and 7B LLMs derived from Gemini technology that beat comparable open models on 11 of 18 text tasks and come with safety assessments.
-
Large Language Models: A Survey
The paper surveys key large language models, their training methods, datasets, evaluation benchmarks, and future research directions in the field.