Xinpeng WangI am an incoming Assistant Professor/Faculty Fellow at NYU CDS and a final-year PhD student at LMU Munich. My supervisor is Prof. Barbara Plank . I was previously a visiting researcher at New York University advised by Prof. He He . Previously, I completed my M.Sc. degree in Robotics, Cognition, Intelligence at Technical University of Munich.
|
|
Research |
|
Is It Thinking or Cheating? Detecting Implicit Reward Hacking by Measuring Reasoning EffortXinpeng Wang*, Nitish Joshi*, Barbara Plank, Rico Angell, He He ICLR, 2026 arxiv / |
|
Refusal Direction is Universal Across Safety-Aligned LanguagesXinpeng Wang*, Mingyang Wang*, Yihong Liu*, Hinrich Schuetze, Barbara Plank NeurIPS, 2025 arxiv / |
|
Surgical, Cheap, and Flexible: Mitigating False Refusal in Language Models via Single Vector AblationXinpeng Wang, Chengzhi Hu, Paul Röttger, Barbara Plank ICLR, 2025 arxiv / code / |
|
Look at the Text: Instruction-Tuned Language Models are More Robust Multiple Choice Selectors than You ThinkXinpeng Wang, Chengzhi Hu, Bolei Ma, Paul Röttger, Barbara Plank COLM, 2024 arxiv / code / |
|
"My Answer is C": First-Token Probabilities Do Not Match Text Answers in Instruction-Tuned Language ModelsXinpeng Wang, Bolei Ma, Chengzhi Hu, Leon Weber-Genzel, Paul Röttger, Frauke Kreuter, Dirk Hovy, Barbara Plank ACL Findings, 2024 arxiv / code / |
|
How to Distill your BERT: An Empirical Study on the Impact of Weight Initialisation and Distillation ObjectivesXinpeng Wang, Leonie Weissweiler, Hinrich Schütze, Barbara Plank ACL, 2023 arxiv / code / |
|
Sceneformer: Indoor Scene Generation with TransformersXinpeng Wang, Chandan Yeshwanth, Matthias Nießner 3DV, 2021 arxiv / video / code / |
|
Design and source code from Jon Barron's website |