6 papers
Projecting Out the Malice: A Global Subspace Approach to LLM Detoxification
Zenghao Duan, Zhiyi Yin, Zhichao Shi +8
Large language models (LLMs) exhibit exceptional performance but pose inherent risks of generating toxic content, restricting their safe deployment. While traditional methods (e.g.…
Circular Reasoning: Understanding Self-Reinforcing Loops in Large Reasoning Models
Zenghao Duan, Liang Pang, Zihao Wei +6
Despite the success of test-time scaling, Large Reasoning Models (LRMs) frequently encounter repetitive loops that lead to computational waste and inference failure. In this paper,…
LLM Latent Reasoning as Chain of Superposition
Jingcheng Deng, Liang Pang, Zihao Wei +6
Latent reasoning offers a computation-efficient alternative to Chain-of-Thought but often suffers from performance degradation due to distributional misalignment and ambiguous chai…
The Evolution of Thought: Tracking LLM Overthinking via Reasoning Dynamics Analysis
Zihao Wei, Liang Pang, Jiahao Liu +7
Test-time scaling via explicit reasoning trajectories significantly boosts large language model (LLM) performance but often triggers overthinking. To explore this, we analyze reaso…
RLKD: Distilling LLMs' Reasoning via Reinforcement Learning
Shicheng Xu, Liang Pang, Yunchang Zhu +6
Distilling reasoning paths from teacher to student models via supervised fine-tuning (SFT) provides a shortcut for improving the reasoning ability of smaller Large Language Models…
Following the Autoregressive Nature of LLM Embeddings via Compression and Alignment
Jingcheng Deng, Zhongtao Jiang, Liang Pang +5
A new trend uses LLMs as dense text encoders via contrastive learning. However, since LLM embeddings predict the probability distribution of the next token, they are inherently gen…