Scholar
Zachary Coalson
Google Scholar ID: AFAZZgkAAAAJ
PhD Student of Computer Science, Oregon State University
Citations & Impact
Citations
12
H-index
2
i10-index
0
Publications
7
Co-authors
1
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Paper 'Using influence functions to reduce language model toxicity' accepted to NeurIPS 2025
- Paper 'Characterizing the resilience of LLM inference to random bit-wise faults' accepted to SC 2025
- Paper 'Improving the efficiency of VLN' accepted to ICCV 2025
- Paper 'Jailbreaking large language models with adversarial bit-flips' posted on arXiv
- Paper 'Poisoning attacks against neural architecture search' posted on arXiv
- First publication 'Auditing multi-exit language models to adversarial slowdown' accepted to NeurIPS 2023
- Awarded 2025 GEM Fellowship
- Recipient of 2025 NSF GRFP
- Received ARCS Foundation Oregon Scholar Award (2023)
Background
- First-year PhD student in Computer Science at Oregon State University
- Researching trustworthy and socially responsible machine learning under the supervision of Professor Sanghyun Hong
- Member of the Trustworthy and Responsible AI Lab (TRUE)
- Aims to audit and improve the robustness of ML systems against adversarial threats, misuse, and undesirable behaviors
- Currently focused on improving trustworthiness of large language models, e.g., reducing toxicity and studying new jailbreaking techniques
