The Research Desk.

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

Last Updated: May 2, 2026, 9:33 AM PT

X.com Research Buzz

The AI Layoff Trap
X.com
17395

The AI Layoff Trap

Brett Hemenway Falk, Gerry Tsoukalas

The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity
X.com
6522

The Illusion of Thinking: Understanding the Strengths and Limitations of Reasoning Models via the Lens of Problem Complexity

Parshin Shojaee, Iman Mirzadeh, Keivan Alizadeh +3 more

#reasoning

AlphaXiv Trending

Incompressible Knowledge Probes: Estimating Black-Box LLM Parameter Counts via Factual Capacity
AlphaXiv
143

Incompressible Knowledge Probes: Estimating Black-Box LLM Parameter Counts via Factual Capacity

Bojie Li

#nlp#alphaxiv
Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation
AlphaXiv
97

Tuna-2: Pixel Embeddings Beat Vision Encoders for Multimodal Understanding and Generation

Zhiheng Liu, Weiming Ren, Xiaoke Huang

#computer-vision#retrieval#multimodal#alphaxiv
A Milestone in Formalization: The Sphere Packing Problem in Dimension 8
AlphaXiv
78

A Milestone in Formalization: The Sphere Packing Problem in Dimension 8

Sidharth Hariharan, Christopher Birkbeck, Seewoo Lee

#alphaxiv
Recursive Multi-Agent Systems
AlphaXiv
67

Recursive Multi-Agent Systems

Xiyuan Yang, Jiaru Zou, Rui Pan

#reinforcement-learning#alphaxiv
Thinking with Visual Primitives
AlphaXiv
51

Thinking with Visual Primitives

Ruijie Lu, Yiyang Ma, Xiaokang Chen

#computer-vision#alphaxiv
Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses
AlphaXiv
47

Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses

Jiahang Lin, Shichun Liu, Chengjun Pan

#reinforcement-learning#reasoning#alphaxiv

HuggingFace Daily Papers

Nemotron 3 Nano Omni: Efficient and Open Multimodal Intelligence
HuggingFace
14

Nemotron 3 Nano Omni: Efficient and Open Multimodal Intelligence

NVIDIA, Amala Sanjay Deshmukh, Kateryna Chumachenko +2 more

#efficiency#multimodal
Step-level Optimization for Efficient Computer-use Agents
HuggingFace
9

Step-level Optimization for Efficient Computer-use Agents

Jinbiao Wei, Kangqi Ni, Yilun Zhao +2 more

#reinforcement-learning#efficiency#yale-nlp
Learning from Noisy Preferences: A Semi-Supervised Learning Approach to Direct Preference Optimization
HuggingFace
3

Learning from Noisy Preferences: A Semi-Supervised Learning Approach to Direct Preference Optimization

Xinxin Liu, Ming Li, Zonglin Lyu +2 more

#efficiency
ViPO: Visual Preference Optimization at Scale
HuggingFace
1

ViPO: Visual Preference Optimization at Scale

Ming Li, Jie Wu, Justin Cui +2 more

#computer-vision#efficiency
FlashRT: Towards Computationally and Memory Efficient Red-Teaming for Prompt Injection and Knowledge Corruption
HuggingFace
0

FlashRT: Towards Computationally and Memory Efficient Red-Teaming for Prompt Injection and Knowledge Corruption

Yanting Wang, Chenlong Yin, Ying Chen +1 more

#nlp#efficiency#wang-yanting
Safety Drift After Fine-Tuning: Evidence from High-Stakes Domains
HuggingFace
0

Safety Drift After Fine-Tuning: Evidence from High-Stakes Domains

Emaan Bilal Khan, Amy Winecoff, Miranda Bogen +1 more

#machine-learning#safety-alignment