Scholar
Xin Chen
Google Scholar ID: KVMYX5QAAAAJ
ETH Zurich
Citations & Impact
Citations
1,074
H-index
7
i10-index
7
Publications
9
Co-authors
5
list available
Resume (English only)
Academic Achievements
- Published several papers, including:
- - 'Learning Safety Constraints for Large Language Models' (ICML2025 Spotlight)
- - 'Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback' (TMLR, Outstanding Paper Finalist)
- - 'Learning Safety Constraints from Demonstrations with Unknown Rewards' (AISTATS2024)
- - 'Arch-Graph: Acyclic Architecture Relation Predictor for Task-Transferable Neural Architecture Search' (CVPR2022)
- - 'An Empirical Investigation of Representation Learning for Imitation' (NeurIPS2021)
- - 'Exploring Geometry-aware Contrast and Clustering Harmonization for Self-supervised 3D Object Detection' (ICCV2021)
- - 'TransNAS-Bench-101: Improving transferability and Generalizability of Cross-Task Neural Architecture Search' (CVPR2021)
- - 'CATCH: Context-based Meta Reinforcement Learning for Transferrable Architecture Search' (ECCV2020)
Research Experience
- Helps organize the International Dialogues on AI Safety.
Education
- Graduated from The University of Hong Kong; spent time at UC Berkeley (Center for Human-Compatible AI), Stanford University, and Columbia University before ETH.
Background
- Currently a Direct PhD student at ETH Zurich, jointly supervised by Prof. Andreas Krause and Prof. Florian Tramèr. Passionate about developing AI solutions that are beneficial to society and align with human values. Research interests include understanding the science of LLM Alignment, learning human values, and reward hacking.
Miscellany
- Supported by the Open Phil AI Fellowship and the Vitalik Buterin PhD Fellowship; donates a considerable portion of her income to the most effective charities every year; aims to read 50 books a year.
