Zachary Coalson

11 papers A* 1A 1Journal 9
YearRankTypeTitle / Venue / Authors
2026 J jnl
Asking Forever: Universal Activations Behind Turn Amplification in Conversational LLMs.
CoRR
Zachary Coalson, Bo Fang, Sanghyun Hong
2026 J jnl
Discovering Universal Activation Directions for PII Leakage in Language Models.
CoRR
Leo Marchyok, Zachary Coalson, Sungho Keum, Sooel Son, Sanghyun Hong
2026 J jnl
Fail-Closed Alignment for Large Language Models.
CoRR
Zachary Coalson, Beth Sohler, Aiden Gabriel, Sanghyun Hong
2025 A conf
Demystifying the Resilience of Large Language Model Inference: An End-to-End Perspective.
SC
Yu Sun, Zachary Coalson, Shiyang Chen, Hang Liu, Zhao Zhang, Sanghyun Hong, Bo Fang, Lishan Yang
2025 J jnl
Hard Work Does Not Always Pay Off: On the Robustness of NAS to Data Poisoning.
Trans. Mach. Learn. Res.
Zachary Coalson, Huazheng Wang, Qingyun Wu, Sanghyun Hong
2025 J jnl
Harnessing Input-Adaptive Inference for Efficient VLN.
CoRR
Dongwoo Kang, Akhil Perincherry, Zachary Coalson, Aiden Gabriel, Stefan Lee, Sanghyun Hong
2025 J jnl
IF-GUIDE: Influence Function-Guided Detoxification of LLMs.
CoRR
Zachary Coalson, Juhan Bae, Nicholas Carlini, Sanghyun Hong
2024 J jnl
Hard Work Does Not Always Pay Off: Poisoning Attacks on Neural Architecture Search.
CoRR
Zachary Coalson, Huazheng Wang, Qingyun Wu, Sanghyun Hong
2024 J jnl
PrisonBreak: Jailbreaking Large Language Models with Fewer Than Twenty-Five Targeted Bit-flips.
CoRR
Zachary Coalson, Jeonghyun Woo, Shiyang Chen, Yu Sun, Lishan Yang, Prashant J. Nair, Bo Fang, Sanghyun Hong
2023 A* conf
BERT Lost Patience Won't Be Robust to Adversarial Slowdown.
NeurIPS
Zachary Coalson, Gabriel Ritter, Rakesh Bobba, Sanghyun Hong
2023 J jnl
BERT Lost Patience Won't Be Robust to Adversarial Slowdown.
CoRR
Zachary Coalson, Gabriel Ritter, Rakesh Bobba, Sanghyun Hong