Distillation Beats Zero-RL: A Simpler Path to Smarter Reasoning? This paper delivers a surprising—and important—result: simple distillation from a stronger model can outperform full-blown reinforcement learning on small models, even with far fewer data and less compute. Key
@jiqizhixin
-
Learning Reasoning Without External Rewards in AI Systems
By
–
Learning to Reason without External Rewards
Paper: https://
arxiv.org/pdf/2505.19590
.pdf
…
Code: https://
github.com/sunblaze-ucb/I
ntuitor
… -
Long-Context State-Space Video World Models Paper
By
–
Long-Context State-Space World Models
Paper: https://
arxiv.org/pdf/2505.20171
.pdf
… -
State-Space Diffusion Models Advance Video World Modeling Memory
By
–
State-Space Meets Diffusion: A New Era for World Models This paper tackles a core bottleneck in video-based world modeling: long-term memory. While video diffusion models are great at short-term frame prediction, their memory fades fast—especially when modeling long
-
AI-Researcher: Autonomous Scientific Innovation System
By
–
AI-Researcher: Autonomous Scientific Innovation
Paper: https://
arxiv.org/pdf/2505.18705
.pdf
…
Code: https://
github.com/HKUDS/AI-Resea
rcher
… -
AI-Researcher: Autonomous System for Scientific Discovery Pipeline
By
–
AI-Researcher: Autonomous Scientific Innovation This paper, from HKU, introduces AI-Researcher, a fully autonomous system that runs the entire research pipeline—yes, from reading papers to writing them. It's the clearest step yet toward AI-led scientific discovery.
-
LLMs and Humans Trade Compression for Meaning Analysis
By
–
From Tokens to Thoughts: How LLMs and Humans Trade Compression for Meaning
Paper: https://
arxiv.org/pdf/2505.17117
v1.pdf
… -
How LLMs Organize Concepts Differently from Human Brains
By
–
Do LLMs Think Like Humans? New Study Says Not Quite This paper dives into a fascinating question:
Do Large Language Models organize concepts the way humans do?
Spoiler: They don’t—at least not yet. Humans use semantic compression—they categorize by balancing expressive -
Superplatforms Attack AI Agents to Defend Digital Monopoly Control
By
–
Superplatforms have to attack AI agents to defend their centralized control of digital traffic entrance. Why? Superplatforms—those giants that bundle countless apps and services into one ecosystem—have long thrived by monopolizing user attention via ads and algorithms. But now,
-
Andrew Ng’s Career Advice: Reading Research Papers Effectively
By
–
This blog summarizes the career advice/reading research papers lecture in the CS230 Deep learning course by Stanford University on YouTube, and includes advice from Andrew Ng on how to read research papers. https://
kdnuggets.com/2019/09/advice
-building-machine-learning-career-research-papers-andrew-ng.html
…
Good stuff never ages.