Unverified50% confidenceFactExact time
Deletion Tuning can match the performance of full fine-tuning under all three post-training paradigms: reinforcement learning with verifiable rewards (RLVR), supervised fine-tuning (SFT) and on-policy distillation.
1
Sources
50%
Confidence
Long-term
Relevance
11/10/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAcross the three post-training paradigms of RLVR, SFT, and on-policy distillation, zeroing out less than 0.05% of the weights in the pretrained model matches the performance of full fine-tuning at scale.79% similarUnverifiedDeletion Tuning is especially suitable for scenarios that require frequent task switching or continual learning.73% similarVerified从预训练检查点到可商用产品需经历监督式指令微调(SFT)和基于人类反馈的强化学习(RLHF)两个后训练阶段72% similarUnverified后训练通常涵盖监督微调(SFT)、强化学习对齐(RLHF/RLAIF)以及领域适配微调,视觉推理模型还涉及视觉指令微调71% similarUnverifiedFine-tuning methods include supervised fine-tuning and reinforcement with a grader.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1157636API
curl https://kongchang.com/api/v1/knowledge/claims/1157636MCP
get_claim(id=1157636)