白寅岐

2026

Adversarial Latent Embedding Repair for LLM Continual Learning

Xilin Xia, Xialiang Tong, Jie Wang, Chi Ma, Shengxue Li, 白寅岐 +5

ICML 2026poster

Research on continual learning for LLMs seeks to acquire new skills without catastrophic forgetting of established prior knowledge. However, domain-specific fine-tuning still triggers severe, long-tailed forgetting issues even under narrow updates, particularly when the pre-training data is inaccess…

Cited by 0 Source ↗Scholar ↗

2026

D-ARL: A Distribution-Matched Asynchronous Reinforcement Learning Framework for Language Reasoning

白寅岐, Xialiang Tong, Jie Wang, Hongyu Liu, Longdi Pan, Jiashuo Li +4

ICML 2026poster

Asynchronous reinforcement learning (RL) has shown notable success in accelerating the post-training of large language models (LLMs). However, its decoupled data generation and training paradigm introduces a fundamental distributional mismatch between data generated by stale behavior policies and cu…

Cited by 0 Source ↗Scholar ↗

白 寅岐

Adversarial Latent Embedding Repair for LLM Continual Learning

D-ARL: A Distribution-Matched Asynchronous Reinforcement Learning Framework for Language Reasoning