11 papers
Generalization Dynamics of LM Pre-training
Jiaxin Wen, Zhengxuan Wu, Dawn Song +1
People typically assume that LMs stably mature from pattern-matching parrots to generalizable intelligence during pre-training. We build a toy eval suite and show this mental model…
The 2026 Singapore Consensus on Global AI Safety Research Priorities
Stephen Casper, Oskar Galeev, Yoshua Bengio +117
Frontier AI capabilities and autonomy are advancing rapidly. A growing number of real-world incidents make a trusted AI ecosystem essential to embracing AI with confidence. The 202…
Strategy Executability in Mathematical Reasoning: Leveraging Human-Model Differences for Effective Guidance
Weida Liang, Yiyou Sun, Shuyuan Nan +3
Example-based guidance is widely used to improve mathematical reasoning at inference time, yet its effectiveness is highly unstable across problems and models-even when the guidanc…
How and Why LLMs Generalize: A Fine-Grained Analysis of LLM Reasoning from Cognitive Behaviors to Low-Level Patterns
Haoyue Bai, Yiyou Sun, Wenjie Hu +5
Large Language Models (LLMs) display strikingly different generalization behaviors: supervised fine-tuning (SFT) often narrows capability, whereas reinforcement-learning (RL) tunin…
RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?
Yiyou Sun, Yuhan Cao, Pohao Huang +4
It remains an open question whether LLMs can acquire or generalize genuinely new reasoning strategies, beyond the sharpened skills encoded in their parameters during pre-training o…
Can Aha Moments Be Fake? Towards Quantifying Decorative and True Thinking in Chain-of-Thought
Jiachen Zhao, Yiyou Sun, Weiyan Shi +1
Large language models can generate long chain-of-thought (CoT) reasoning, yet prior work suggests that CoT can be post-hoc rationalization rather than a faithful reflection of the…
MIRAGE-Bench: LLM Agent is Hallucinating and Where to Find Them
Weichen Zhang, Yiyou Sun, Pohao Huang +3
Hallucinations pose critical risks for large language model (LLM)-based agents, often manifesting as hallucinative actions resulting from fabricated or misinterpreted information w…
The Singapore Consensus on Global AI Safety Research Priorities
Yoshua Bengio, Tegan Maharaj, Luke Ong +84
Rapidly improving AI capabilities and autonomy hold significant promise of transformation, but are also driving vigorous debate on how to ensure that AI is safe, i.e., trustworthy,…
OMEGA: Can LLMs Reason Outside the Box in Math? Evaluating Exploratory, Compositional, and Transformative Generalization
Yiyou Sun, Shawn Hu, Georgia Zhou +4
Recent large-scale language models (LLMs) with long Chain-of-Thought reasoning-such as DeepSeek-R1-have achieved impressive results on Olympiad-level mathematics benchmarks. Howeve…
Why and How LLMs Hallucinate: Connecting the Dots with Subsequence Associations
Yiyou Sun, Yu Gai, Lijie Chen +3
Large language models (LLMs) frequently generate hallucinations-content that deviates from factual accuracy or provided context-posing challenges for diagnosis due to the complex i…
Climbing the Ladder of Reasoning: What LLMs Can-and Still Can't-Solve after SFT?
Yiyou Sun, Georgia Zhou, Haoyue Bai +4
Recent supervised fine-tuning (SFT) approaches have significantly improved language models' performance on mathematical reasoning tasks, even when models are trained at a small sca…