5 papers · 1 filter
Not All Rollouts Are Worth Learning: On Trajectory Valuation for Post-Training Reinforcement Learning
Xuesong Jia, Ziao Yang, Zhanhe Huang +1
We consider the problem of trajectory valuation in reinforcement learning: how to identify and mitigate detrimental trajectories during online training. Unlike classification, wher…
Validation-Induced Shapley Shifts: How Validation Structure Distorts Data Valuation
Yinan Shen, Ziao Yang, Hongfu Liu
Shapley values are widely used to attribute value to training data based on their marginal contribution to performance on a validation set. Existing practice often assumes these va…
Layer-Aware Influence for Online Data Valuation Estimation
Ziao Yang, Longbo Huang, Hongfu Liu
Data-centric learning emphasizes curating high-quality training samples to boost performance rather than designing new architectures. A central problem is to estimate the influence…
On the Inflation of KNN-Shapley Value
Ziao Yang, Han Yue, Jian Chen +1
Shapley value-based data valuation methods, originating from cooperative game theory, quantify the usefulness of each individual sample by considering its contribution to all possi…
Revisit, Extend, and Enhance Hessian-Free Influence Functions
Ziao Yang, Han Yue, Jian Chen +1
Influence functions serve as crucial tools for assessing sample influence in model interpretation, subset training set selection, noisy label detection, and more. By employing the…