影响力指数

93.77/100

前 0.3%

全站排名 #217

发表论文26 篇

平均评分5.9

年均产出8.7 篇/年

Steven Wu

Assistant Professor@Carnegie Mellon University·美国·OpenReview

研究方向

Reinforcement Learning · Differential Privacy · Algorithmic Fairness · Algorithmic Economics · Online Learning

Dimension-Free Decision Calibration for Nonlinear Loss Functions

ICLR 2026Poster

Doubly-Robust LLM-as-a-Judge: Externally Valid Estimation with Imperfect Personas

ICLR 2026Poster

All Roads Lead to Likelihood: The Value of Reinforcement Learning in Fine-Tuning

ICLR 2026Poster

From Curiosity to Caution: Mitigating Reward Hacking for Best-of-$N$ with Pessimism

ICLR 2026Poster

Persuasive Prediction via Decision Calibration

ICLR 2026Rejected

Nearly-Optimal Bandit Learning in Stackelberg Games with Side Information

ICLR 2026Poster

Optimizing Canaries for Privacy Auditing with Metagradient Descent

ICLR 2026Poster

Membership Inference Attacks for Unseen Classes

ICLR 2026Rejected

BLUR: A Benchmark for LLM Unlearning Robust to Forget-Retain Overlap

ICLR 2026Rejected

Validating LLM-as-a-Judge Systems under Rating Indeterminacy

NeurIPS 2025Poster

Discretization-free Multicalibration through Loss Minimization over Tree Ensembles

NeurIPS 2025Poster

Utility-Directed Conformal Prediction: A Decision-Aware Framework for Actionable Uncertainty Quantification

ICLR 2025Poster

Unlearning or Obfuscating? Jogging the Memory of Unlearned LLMs via Benign Relearning

ICLR 2025Poster

Leveraging Model Guidance to Extract Training Data from Personalized Diffusion Models

ICML 2025Poster

Kandinsky Conformal Prediction: Beyond Class- and Covariate-Conditional Coverage

ICML 2025Poster

Unlearned but Not Forgotten: Data Extraction after Exact Unlearning in LLM

NeurIPS 2025Poster

Reconciling Model Multiplicity for Downstream Decision Making

ICLR 2025Poster

Revealing the Unseen: Guiding Personalized Diffusion Models to Expose Training Data

ICLR 2025Rejected

合作者 (20)

Pratiksha Thaker