← Search

Ding Cao

2 accepted papers

2026

On Predictability of Reinforcement Learning Dynamics for Large Language Models

ICLR 2026poster

Recent advances in reasoning capabilities of large language models (LLMs) are largely driven by reinforcement learning (RL), yet the underlying parameter dynamics during RL training remain poorly understood. This work identifies two fundamental properties of RL-induced parameter updates in LLMs: (1)…

Cited by 0SourcecodeScholar
2026

On the Superimposed Noise Accumulation Problem in Sequential Knowledge Editing of Large Language Models

AAAI 2026technical

Sequential knowledge editing techniques aim to continuously update knowledge in large language models at low cost, preventing models from generating outdated or incorrect information. However, existing sequential editing methods suffer from a significant decline in editing success rates after long-t

Cited by 0SourcePDFScholar