跳到正文
arXiv cs.AI· Mingda Zhang, Yuanhao Dong, Qiang Huang, Rui Qian, Yanjin Li, Zhenyu Liu, Xiao Huang, Zihao Yang, Jian Wang, Pinjia He, Xiaoying Tang·· 10 小时前AI 评分38

BaseRSI:通过免训练内部化 Harness 实现基础视觉语言模型的递归自我改进

BaseRSI: Recursive Self-Improvement of a Base Vision-Language Model via Training-Free Harness Internalization

AI 导读

研究者提出 BaseRSI,一种让基础视觉语言模型递归自我改进的新范式:模型先改进自身 harness,再通过免训练编辑把 harness 带来的增益写入自身权重。

正文 · 原文

Authors:Mingda Zhang, Yuanhao Dong, Qiang Huang, Rui Qian, Yanjin Li, Zhenyu Liu, Xiao Huang, Zihao Yang, Jian Wang, Pinjia He, Xiaoying Tang

View PDF HTML (experimental)

Abstract:Recursive self-improvement (RSI) is moving from the harness of prompts and tools around a base model toward the model itself. However, existing RSI still faces key challenges, including harness-bound gains that stay outside the base model, black-box supervision that reads only the model's outputs, and imprecise edits whose site and strength are fixed in advance. To address these challenges, we propose BaseRSI, a new paradigm of Recursive Harness Internalization: the base model improves its harness, then writes what the harness adds into its own weights through training-free edits, round after round on its own responses. To support this paradigm, we introduce ReGap, a mechanism that supervises every transformer block with the high-dimensional activation difference that the harness makes at that block alone, and updates the block's weights along this signal in a low-dimensional subspace. Building on this mechanism, we further propose ReKey, an algorithm that writes this signal by regression into a few key-value neurons of every block, and measures the site and strength of each edit on the model's own responses. Moreover, the base model fills every role of this recursion with forward passes alone, and its edits, written at different sites, compose in one checkpoint without refitting. Experimental results on ten visual reasoning benchmarks show that BaseRSI consistently improves a base vision-language model both in distribution and on held-out benchmarks. Our code is available at this https URL.
Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.24578 [cs.CV]
  (or arXiv:2611.24578v1 [cs.CV] for this version)
  https://doi.org/10.48550/arXiv.2611.24578

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Mingda Zhang [view email]
[v1] Wed, 18 Nov 2026 15:10:11 UTC (26,825 KB)

来源:arXiv cs.AI · arxiv.org