6 papers
Dynamic Rollout Editing for Reducing Overthinking in RL-Trained Reasoning Models
Zihao Wei, Wenjie Shi, Liang Pang +8
Long-form chain-of-thought reasoning can improve LLM performance on complex tasks, but models often continue generating unnecessary reasoning after a correct answer has emerged. We…
BaseCal: Unsupervised Confidence Calibration via Base Model Signals
Hexiang Tan, Wanli Yang, Junwei Zhang +7
Reliable confidence is essential for trusting the outputs of LLMs, yet widely deployed post-trained LLMs (PoLLMs) typically compromise this trust with severe overconfidence. In con…
Beyond Reasoning: Reinforcement Learning Unlocks Parametric Knowledge in LLMs
Wanli Yang, Hongyu Zang, Junwei Zhang +5
Reinforcement learning (RL) has achieved remarkable success in LLM reasoning, but whether it can also improve direct recall of parametric knowledge remains an open question. We stu…
Fine-tuning Done Right in Model Editing
Wanli Yang, Rui Tang, Hongyu Zang +6
Fine-tuning, a foundational method for adapting large language models, has long been considered ineffective for model editing. Here, we challenge this belief, arguing that the repo…
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…
Too Consistent to Detect: A Study of Self-Consistent Errors in LLMs
Hexiang Tan, Fei Sun, Sha Liu +8
As large language models (LLMs) often generate plausible but incorrect content, error detection has become increasingly critical to ensure truthfulness. However, existing detection…