Meta Research

company

https://research.facebook.com/

facebookresearch

Activity Feed Request to join this org

AI & ML interests

None defined yet.

Recent Activity

QuantaSparkLabs updated a model about 3 hours ago

metaresearch/PapersRAG-1.5B

Gaetan10 submitted a paper about 23 hours ago

Distilling LLM Feedback for Lean Theorem Proving

Benjamin-eecs authored a paper 4 days ago

Agents' Last Exam

View all activity

Papers

OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification

Distilling LLM Feedback for Lean Theorem Proving

View all Papers

metaresearch 's papers 28

Submitted by

Yuhang Zhou

OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification

metaresearch

Submitted by

N

Distilling LLM Feedback for Lean Theorem Proving

metaresearch

2

DUEL: Adversarial Self-Play for Multimodal Reasoning

metaresearch

Submitted by

Feiyan Zhou

WavFlow: Audio Generation in Waveform Space

metaresearch

TARo: Token-level Adaptive Routing for LLM Test-time Alignment

metaresearch

Submitted by

Ruizhong Qiu

ReMix: Reinforcement routing for mixtures of LoRAs in LLM finetuning

metaresearch

4

Submitted by

taesiri

Towards a Neural Debugger for Python

metaresearch

Submitted by

taesiri

APRES: An Agentic Paper Revision and Evaluation System

metaresearch

Submitted by

taesiri

Gaia2: Benchmarking LLM Agents on Dynamic and Asynchronous Environments

metaresearch

Submitted by

taesiri

Likelihood-Based Reward Designs for General LLM Reasoning

metaresearch

Submitted by

Lisa Alazraki

Scaling Small Agents Through Strategy Auctions

metaresearch

3

Submitted by

taesiri

Self-Improving Pretraining: using post-trained models to pretrain better models

metaresearch

Submitted by

Lin G

Paying Less Generalization Tax: A Cross-Domain Generalization Study of RL Training for LLM Agents

metaresearch

Submitted by

taesiri

Agentic Very Long Video Understanding

metaresearch

Submitted by

taesiri

Action100M: A Large-scale Video Action Dataset

metaresearch

Submitted by

Gang Liao

KernelEvolve: Scaling Agentic Kernel Coding for Heterogeneous AI Accelerators at Meta

metaresearch

3

Submitted by

taesiri

Exploring MLLM-Diffusion Information Transfer with MetaCanvas

metaresearch

Submitted by

taesiri

Confucius Code Agent: An Open-sourced AI Software Engineer at Industrial Scale

metaresearch

Submitted by

Jason Weston

AI & Human Co-Improvement for Safer Co-Superintelligence

metaresearch

Submitted by

taesiri

Self-Improving VLM Judges Without Human Annotations

metaresearch

Submitted by

Yunkai Zhang

Don't Waste It: Guiding Generative Recommenders with Structured Human Priors via Multi-head Decoding

metaresearch

Submitted by

taesiri

Scaling Agent Learning via Experience Synthesis

metaresearch

Submitted by

taesiri

Embody 3D: A Large-scale Multimodal Motion and Behavior Dataset

metaresearch

Submitted by

Chenyu Wang

SPG: Sandwiched Policy Gradient for Masked Diffusion Language Models

metaresearch

Submitted by

taesiri

Agent Learning via Early Experience

metaresearch

Submitted by

Mufei Li

Haystack Engineering: Context Engineering for Heterogeneous and Agentic Long-Context Evaluation

metaresearch

MCGrad: Multicalibration at Web Scale

metaresearch

S'MoRE: Structural Mixture of Residual Experts for LLM Fine-tuning

metaresearch