Xinpeng Wang

22 papers A* 2Journal 14Unranked 6
YearRankTypeTitle / Venue / Authors
2025 — conf
Algorithmic Fidelity of Large Language Models in Generating Synthetic German Public Opinions: A Case Study.
ACL (1)
Bolei Ma, Berk Yoztyurk, Anna-Carolina Haensch, Xinpeng Wang, Markus Herklotz, Frauke Kreuter, Barbara Plank, Matthias Aßenmacher
2025 J jnl
Is It Thinking or Cheating? Detecting Implicit Reward Hacking by Measuring Reasoning Effort.
CoRR
Xinpeng Wang, Nitish Joshi, Barbara Plank, Rico Angell, He He
2025 J jnl
Refusal Direction is Universal Across Safety-Aligned Languages.
CoRR
Xinpeng Wang, Mingyang Wang, Yihong Liu, Hinrich Schütze, Barbara Plank
2025 A* conf
Surgical, Cheap, and Flexible: Mitigating False Refusal in Language Models via Single Vector Ablation.
ICLR
Xinpeng Wang, Chengzhi Hu, Paul Röttger, Barbara Plank
2025 J jnl
Think Before Refusal : Triggering Safety Reflection in LLMs to Mitigate False Refusal Behavior.
CoRR
Shengyun Si, Xinpeng Wang, Guangyao Zhai, Nassir Navab, Barbara Plank
2024 — conf
"My Answer is C": First-Token Probabilities Do Not Match Text Answers in Instruction-Tuned Language Models.
ACL (Findings)
Xinpeng Wang, Bolei Ma, Chengzhi Hu, Leon Weber-Genzel, Paul Röttger, Frauke Kreuter, Dirk Hovy, Barbara Plank
2024 J jnl
"My Answer is C": First-Token Probabilities Do Not Match Text Answers in Instruction-Tuned Language Models.
CoRR
Xinpeng Wang, Bolei Ma, Chengzhi Hu, Leon Weber-Genzel, Paul Röttger, Frauke Kreuter, Dirk Hovy, Barbara Plank
2024 — conf
"Seeing the Big through the Small": Can LLMs Approximate Human Judgment Distributions on NLI from a Few Explanations?
EMNLP (Findings)
Beiduo Chen, Xinpeng Wang, Siyao Peng, Robert Litschko, Anna Korhonen, Barbara Plank
2024 J jnl
"Seeing the Big through the Small": Can LLMs Approximate Human Judgment Distributions on NLI from a Few Explanations?
CoRR
Beiduo Chen, Xinpeng Wang, Siyao Peng, Robert Litschko, Anna Korhonen, Barbara Plank
2024 J jnl
Algorithmic Fidelity of Large Language Models in Generating Synthetic German Public Opinions: A Case Study.
CoRR
Bolei Ma, Berk Yoztyurk, Anna-Carolina Haensch, Xinpeng Wang, Markus Herklotz, Frauke Kreuter, Barbara Plank, Matthias Aßenmacher
2024 J jnl
FinerCut: Finer-grained Interpretable Layer Pruning for Large Language Models.
CoRR
Yang Zhang, Yawei Li, Xinpeng Wang, Qianli Shen, Barbara Plank, Bernd Bischl, Mina Rezaei, Kenji Kawaguchi
2024 J jnl
Look at the Text: Instruction-Tuned Language Models are More Robust Multiple Choice Selectors than You Think.
CoRR
Xinpeng Wang, Chengzhi Hu, Bolei Ma, Paul Röttger, Barbara Plank
2024 J jnl
Surgical, Cheap, and Flexible: Mitigating False Refusal in Language Models via Single Vector Ablation.
CoRR
Xinpeng Wang, Chengzhi Hu, Paul Röttger, Barbara Plank
2024 — conf
The Potential and Challenges of Evaluating Attitudes, Opinions, and Values in Large Language Models.
EMNLP (Findings)
Bolei Ma, Xinpeng Wang, Tiancheng Hu, Anna-Carolina Haensch, Michael A. Hedderich, Barbara Plank, Frauke Kreuter
2024 J jnl
The Potential and Challenges of Evaluating Attitudes, Opinions, and Values in Large Language Models.
CoRR
Bolei Ma, Xinpeng Wang, Tiancheng Hu, Anna-Carolina Haensch, Michael A. Hedderich, Barbara Plank, Frauke Kreuter
2024 J jnl
Understanding When Tree of Thoughts Succeeds: Larger Models Excel in Generation, Not Discrimination.
CoRR
Qiqi Chen, Xinpeng Wang, Philipp Mondorf, Michael A. Hedderich, Barbara Plank
2023 A* conf
ACTOR: Active Learning with Annotator-specific Classification Heads to Embrace Human Label Variation.
EMNLP
Xinpeng Wang, Barbara Plank
2023 J jnl
ACTOR: Active Learning with Annotator-specific Classification Heads to Embrace Human Label Variation.
CoRR
Xinpeng Wang, Barbara Plank
2023 — conf
How to Distill your BERT: An Empirical Study on the Impact of Weight Initialisation and Distillation Objectives.
ACL (2)
Xinpeng Wang, Leonie Weissweiler, Hinrich Schütze, Barbara Plank
2023 J jnl
How to Distill your BERT: An Empirical Study on the Impact of Weight Initialisation and Distillation Objectives.
CoRR
Xinpeng Wang, Leonie Weissweiler, Hinrich Schütze, Barbara Plank
2021 — conf
SceneFormer: Indoor Scene Generation with Transformers.
3DV
Xinpeng Wang, Chandan Yeshwanth, Matthias Nießner
2020 J jnl
SceneFormer: Indoor Scene Generation with Transformers.
CoRR
Xinpeng Wang, Chandan Yeshwanth, Matthias Nießner