8 papers
Why Does RL Generalize Better Than SFT? A Data-Centric Perspective on VLM Post-Training
Aojun Lu, Tao Feng, Hangjie Yuan +2
The adaptation of large-scale Vision-Language Models (VLMs) through post-training reveals a pronounced generalization gap: models fine-tuned with Reinforcement Learning (RL) consis…
Adapt before Continual Learning
Aojun Lu, Tao Feng, Hangjie Yuan +2
Continual Learning (CL) seeks to enable neural networks to incrementally acquire new knowledge (plasticity) while retaining existing knowledge (stability). Although pre-trained mod…
C-Flat++: Towards a More Efficient and Powerful Framework for Continual Learning
Wei Li, Hangjie Yuan, Zixiang Zhao +4
Balancing sensitivity to new tasks and stability for retaining past knowledge is crucial in continual learning (CL). Recently, sharpness-aware minimization has proven effective in…
Achieving Deep Continual Learning via Evolution
Aojun Lu, Junchao Ke, Chunhui Ding +3
Deep neural networks, despite their remarkable success, remain fundamentally limited in their ability to perform Continual Learning (CL). While most current methods aim to enhance…
Runtime Analysis of Evolutionary NAS for Multiclass Classification
Zeqiong Lv, Chao Qian, Yun Liu +2
Evolutionary neural architecture search (ENAS) is a key part of evolutionary machine learning, which commonly utilizes evolutionary algorithms (EAs) to automatically design high-pe…
Loss Functions for Predictor-based Neural Architecture Search
Han Ji, Yuqi Feng, Jiahao Fan +1
Evaluation is a critical but costly procedure in neural architecture search (NAS). Performance predictors have been widely adopted to reduce evaluation costs by directly estimating…