The Research Desk.

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

Last Brew Time: May 7, 2026, 9:41 AM PT

X.com Research Buzz

NLP
Who's in Charge? Disempowerment Patterns in Real-World LLM Usage
X.com
4735

Who's in Charge? Disempowerment Patterns in Real-World LLM Usage

Mrinank Sharma, Miles McCain, Raymond Douglas, David Duvenaud

Anthropic, ACS Research Group, University of Toronto

Reinforcement Learning
Agents of Chaos
X.com
4104

Agents of Chaos

Natalie Shapira, Chris Wendler, Avery Yen, Gabriele Sarti, Koyena Pal, Olivia Floody, Adam Belfki, Alex Loftus, Aditya Ratan Jannali, Nikhil Prakash

Northeastern University, Independent Researcher, Stanford University, University of British Columbia, Harvard University, Hebrew University, Max Planck Institute for Biological Cybernetics, MIT, Tufts University, Carnegie Mellon University, Alter, Technion

NLP
Language models transmit behavioural traits through hidden signals in data
X.com
2101

Language models transmit behavioural traits through hidden signals in data

Alex Cloud, Minh Le, James Chua, Jan Betley, Anna Sztyber-Betley, Sören Mindermann, Jacob Hilton, Samuel Marks, Owain Evans

Anthropic, Truthful AI, Warsaw University of Technology, Oxford Martin AI Governance Initiative, Alignment Research Center, University of California, Berkeley

AlphaXiv Trending

Computer Vision
Thinking with Visual Primitives
AlphaXiv
219

Thinking with Visual Primitives

Ruijie Lu, Yiyang Ma, Xiaokang Chen

Tsinghua University, Peking University

Machine Learning
Let ViT Speak: Generative Language-Image Pre-training
AlphaXiv
97

Let ViT Speak: Generative Language-Image Pre-training

ByteDance, Yan Fang, Mengcheng Lan, Zilong Huang

Beijing Jiaotong University, ByteDance, Nanyang Technological University

Reasoning
MolmoAct2: Action Reasoning Models for Real-world Deployment
AlphaXiv
61

MolmoAct2: Action Reasoning Models for Real-world Deployment

Haoquan Fang, Jiafei Duan, Donovan Clay

Allen Institute for AI, University of Washington, National University of Singapore, University of Pennsylvania, Johns Hopkins University, Amazon, Cortex AI, University of Michigan, University of North Carolina at Chapel Hill

Machine Learning
Model Spec Midtraining: Improving How Alignment Training Generalizes
AlphaXiv
57

Model Spec Midtraining: Improving How Alignment Training Generalizes

Chloe Li, Sara Price, Samuel Marks

Anthropic

Reinforcement Learning
On-Policy Distillation
AlphaXiv
51

On-Policy Distillation

Thinking Machines, Kevin Lu

Thinking Machines Lab

Machine Learning
A Theory of Generalization in Deep Learning
AlphaXiv
46

A Theory of Generalization in Deep Learning

Elon Litman, Gabe Guo

Stanford University

HuggingFace Daily Papers

OpenBMB
MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction
HuggingFace
4

MiniCPM-o 4.5: Towards Real-Time Full-Duplex Omni-Modal Interaction

Junbo Cui, Bokai Xu, Chongyi Wang, Tianyu Yu, Weiyue Sun

OpenBMB

Reinforcement Learning
SWE-WebDevBench: Evaluating Coding Agent Application Platforms as Virtual Software Agencies
HuggingFace
2

SWE-WebDevBench: Evaluating Coding Agent Application Platforms as Virtual Software Agencies

Siddhant Saxena, Nilesh Trivedi, Vinayaka Jyothi

QwikBuild

Reinforcement Learning
CreativityBench: Evaluating Agent Creative Reasoning via Affordance-Based Tool Repurposing
HuggingFace
1

CreativityBench: Evaluating Agent Creative Reasoning via Affordance-Based Tool Repurposing

Cheng Qian, Hyeonjeong Ha, Jiayu Liu, Jeonghwan Kim, Jiateng Liu

University of Illinois at Urbana-Champaign

Safety Alignment
The First Token Knows: Single-Decode Confidence for Hallucination Detection
HuggingFace
1

The First Token Knows: Single-Decode Confidence for Hallucination Detection

Mina Gabriel

Temple University

NLP
When to Think, When to Speak: Learning Disclosure Policies for LLM Reasoning
HuggingFace
1

When to Think, When to Speak: Learning Disclosure Policies for LLM Reasoning

Jiaqi Wei, Xuehang Guo, Pengfei Yu, Xiang Zhang, Wanli Ouyang

Zhejiang University, College of William and Mary, University of Illinois Urbana-Champaign, University of British Columbia, Chinese University of Hong Kong, Fudan University, Stony Brook University

Computer Vision
TT4D: A Pipeline and Dataset for Table Tennis 4D Reconstruction From Monocular Videos
HuggingFace
1

TT4D: A Pipeline and Dataset for Table Tennis 4D Reconstruction From Monocular Videos

Nima Rahmanian, Daniel Kienzle, Thomas Gossard, Dvij Kalaria, Rainer Lienhart

Chair for Machine Learning & Computer Vision