cs.LGNov 7, 2026

The Surprising Effectiveness of Deleting Weights in LLM Post-Training

Authors: Jack Lu, Zhenbang Yang, Mike Lasby, Tejas Pote, Yani Ioannou, Mengye Ren

Organizations: New York University · University of Calgary

Abstract

How little of a pretrained large language model (LLM) has to change for it to acquire new capabilities? We find that zeroing fewer than 0.05% of its weights, with no other modification, matches full fine-tuning (FFT) at scale across three LLM post-training regimes: reinforcement learning with verifiable rewards, supervised fine-tuning, and on-policy distillation. Our method, Deletion Tuning, learns a binary mask over the gate projection of transformer blocks. Two practical advantages follow: the trained mask is hundreds to thousands of times smaller than the corresponding FFT checkpoint, and Deletion Tuning forgets far less, which addresses a central challenge in continual learning. Under reinforcement learning, its prior-task accuracy stays near the base level while FFT, LoRA, and random sparse fine-tuning lose up to 5.4 points. To understand why masking alone suffices, we analyze the geometry of the learned weight changes under reinforcement learning using stable rank, spectrum drift, and principal-weight overlap. Compared with LoRA and random sparse fine-tuning, Deletion Tuning is closest to FFT across the three measures.

Figures & tables

Appendix figures & tables8 assets

Supplementary material from the paper’s appendix.

Appendix

Explore similar work

CardsList
  1. Towards Continual Learning: Mitigating Forgetting With Subspace Tuning

    Oct 15, 2026Arnav Chavan, Kusha Sareen, Rishabh Tiwari +6Continual Learning for LLMsLLM Fine-Tuning

  2. Super Weights in LLMs and the Failure of Selective Training

    Jul 9, 2026Shreyas Subramanian, Adewale Akinfaderin, Akarsha SehwagLLM PruningFine-Tuning

  3. Boosting Large Language Models with Mask Fine-Tuning

    Mar 27, 2025Mingyuan Zhang, Yue Bai, Huan Wang +4LLM Fine-Tuning