3 papers
cs.LG2025
Zeroth-Order Optimization Finds Flat Minima
Liang Zhang, Bingcong Li, Kiran Koshy Thekumparampil +3
Zeroth-order methods are extensively used in machine learning applications where gradients are infeasible or expensive to compute, such as black-box attacks, reinforcement learning…
cs.LG2025
PoLAR: Polar-Decomposed Low-Rank Adapter Representation
Kai Lion, Liang Zhang, Bingcong Li +1
We show that low-rank adaptation of large-scale models suffers from a low stable rank that is well below the linear algebraic rank of the subspace, degrading fine-tuning performanc…
cs.LG2025
Can RLHF be More Efficient with Imperfect Reward Models? A Policy Coverage Perspective
Jiawei Huang, Bingcong Li, Christoph Dann +1
Sample efficiency is critical for online Reinforcement Learning from Human Feedback (RLHF). While existing works investigate sample-efficient online exploration strategies, the pot…