Antidistillation Sampling

Frontier models that generate extended reasoning traces inadvertently produce token sequences that can facilitate model distillation. Recognizing this vulnerability, model owners may seek sampling strategies that limit the effectiveness of distillation without compromising model performance. Antidistillation sampling provides exactly this capability. By strategically modifying a model's next-token probability distribution, antidistillation sampling poisons reasoning traces, rendering them significantly less effective for distillation while preserving the model's utility.

Authors

• Yash Savani
• Asher Trockman
• Zhili Feng
• Yixuan Xu
• Avi Schwarzschild
• Alexander Robey
• Marc Finzi
• Zico Kolter

Keywords

No keywords are indexed for this paper.