The Research Desk.

The most upvoted and starred AI research crossing the community today.

Last Brew Time: May 20, 2026, 7:35 AM PT

X.com Research Buzz

2028: Two scenarios for global AI leadership
X.com
8108

2028: Two scenarios for global AI leadership

Anthropic

AlphaXiv Trending

ELF: Embedded Language Flows
AlphaXiv
293

ELF: Embedded Language Flows

Yiyang Lu, Keya Hu, Linlu Qiu

MIT

Reinforcement Learning
Self-Distilled Agentic Reinforcement Learning
AlphaXiv
195

Self-Distilled Agentic Reinforcement Learning

Zhengxi Lu, Zhiyuan Yao, Zhuowen Han

Zhejiang University, Tsinghua University

Robotics
World Action Models: The Next Frontier in Embodied AI
AlphaXiv
174

World Action Models: The Next Frontier in Embodied AI

Siyin Wang, Junhao Shi, Zhaoyang Fu

VGGT- Ω Ω
AlphaXiv
156

VGGT- Ω Ω

Jianyuan Wang, Minghao Chen, Shangzhan Zhang

NLP
SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Linear Diffusion Transformer
AlphaXiv
125

SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Linear Diffusion Transformer

Haoyi Zhu, Haozhe Liu, Yuyang Zhao

NVIDIA

Reinforcement Learning
FutureSim: Replaying World Events to Evaluate Adaptive Agents
AlphaXiv
110

FutureSim: Replaying World Events to Evaluate Adaptive Agents

Arvindh Arun, Shashwat Goel, Nikhil Chandak

HuggingFace Daily Papers

Efficiency
Active Learners as Efficient PRP Rerankers
HuggingFace
33

Active Learners as Efficient PRP Rerankers

Jeremías Figueiredo Paschmann, Juan Kaplan, Francisco Nattero, Santiago Barron, Juan Wisznia

Universidad de San Andrés

Machine Learning
Overcoming Catastrophic Forgetting in Visual Continual Learning with Reinforcement Fine-Tuning
HuggingFace
5

Overcoming Catastrophic Forgetting in Visual Continual Learning with Reinforcement Fine-Tuning

Meng Lou, Hanzhong Guo, Linwei Chen, Yizhou Yu

NLP
Stage-adaptive Token Selection for Efficient Omni-modal LLMs
HuggingFace
2

Stage-adaptive Token Selection for Efficient Omni-modal LLMs

Zijie Xin, Jie Yang, Ruixiang Zhao, Tianyi Wang, Fengyun Rao

Machine Learning
TideGS: Scalable Training of Over One Billion 3D Gaussian Splatting Primitives via Out-of-Core Optimization
HuggingFace
2

TideGS: Scalable Training of Over One Billion 3D Gaussian Splatting Primitives via Out-of-Core Optimization

Chonghao Zhong, Linfeng Shi, Hua Chen, Tiecheng Sun, Hao Zhao

Sponge Computing Lab at HKUST

NLP
Language-Switching Triggers Take a Latent Detour Through Language Models
HuggingFace
2

Language-Switching Triggers Take a Latent Detour Through Language Models

Francis Kulumba, Wissam Antoun, Théo Lasnier, Benoît Sagot, Djamé Seddah

ALMAnaCH (Inria)

optimize_anything: A Universal API for Optimizing any Text Parameter
HuggingFace
0

optimize_anything: A Universal API for Optimizing any Text Parameter

Lakshya A Agrawal, Donghyun Lee, Shangyin Tan, Wenjie Ma, Karim Elmaaroufi