# The Surprising Effectiveness of Deleting Weights in LLM Post-Training > Source: > Published: 2026-11-11 13:00:05+00:00 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 verifi