[ICLR 2025] Improving Data Efficiency via Curating LLM-Driven Rating Systems

This repository accompanies the paper To Give or Not to Give? The Impacts of Strategically Withheld Recourse published in AISTATS 2025.

[ICLR 2025] FLAT: LLM Unlearning via Loss Adjustment with Only Forget Data

This is an implementation of the paper: Noise-Resilient Point-wise Anomaly Detection in Time Series Using Weak Segment Labels (KDD 25)

[NeurIPS'24] Fairness Without Harm: An Influence-Guided Active Sampling Approach

This repository accompanies the paper Preformative Prediction with Bandit Feedback:Learning through Reparameterization accepted by ICML 2024.

This repository accompanies the paper Model Transferability with Responsive Decision Subjects accepted by ICML 2023.

Identifiability of Label Noise Transition Matrix

Human annotated noisy labels for CIFAR-10 and CIFAR-100. The website of CIFAR-N is available at http://www.noisylabels.com/.