跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 16 天前AI 評分46

探討權重空間編程:能力、泛化與限制

Exploring Weight-Space Programming: Capabilities, Generalization, and Limitations

AI 導讀

研究顯示,Weight-Space Programming(WSP)可透過局部修改神經網絡參數提升推理表現,但跨任務泛化和編輯可靠性並不穩定。研究涵蓋七個基準及 Qwen3.5-0.8B、Qwen3.5-9B;只修改 1,023 個參數,MATH-500 準確率便提升 29.45 個百分點。16 條編輯軌跡中有六條保留原有權重,反映可靠驗證仍具挑戰。

正文

Published on Nov 6

Authors:

,

,

,

,

,

,

Abstract

Weight-Space Programming can improve reasoning performance through small, localized edits to neural network parameters, but its editing reliability and cross-task generalization remain uneven.

Weight-Space Programming (WSP) enables reasoning agents to directly inspect and modify neural network parameters, offering an alternative to gradient-based model adaptation. However, its capabilities, cross-task generalization, and practical limitations remain insufficiently understood. We systematically investigate WSP through 16 editing trajectories across seven benchmarks and two model scales (Qwen3.5-0.8B and Qwen3.5-9B). Our findings reveal three key insights. First, WSP can achieve substantial performance gains through highly localized interventions. On full held-out evaluations, accuracy improves by up to 29.45, 4.42, 36.72, and 17.11 percentage points on MATH-500, GPQA Diamond, IFEval, and SciBench, respectively. Notably, modifying only 1,023 parameters yields a 29.45-point improvement. Second, WSP exhibits promising but uneven cross-task generalization. A Math-edited checkpoint improves GPQA Diamond, IFEval, and SciBench by 3.03, 12.18, and 20.10 points, respectively, without further editing, while leaving SciCode performance unchanged. Third, effective editing remains challenging and inconsistent. Outcomes vary substantially across trajectories, and six of the 16 trajectories retain the original weights when candidate improvements cannot be reliably validated. Together, these results highlight WSP's potential for highly localized model adaptation while revealing important challenges in intervention discovery, reliability, and cross-task transfer. Homepage: https://wsp-scope.github.io/ GitHub: https://github.com/WSP-Scope/WSP-Scope-Project

View arXiv page View PDF Project page GitHub Add to collection

Get this paper in your agent:

hf papers read 2611.09382

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2611.09382 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2611.09382 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2611.09382 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co