English(EN) Interpreting vision transformers via residual replacement model

新模型为理解视觉 Transformer 机制提供可解释的洞察

作者 PulseAugur 编辑部 · [1 个来源] · 2026-10-23 04:00

研究人员开发了一种新颖的残差替换模型来解释视觉 Transformer (ViTs) 的内部工作原理。该模型分析所有层的特征，揭示 ViTs 如何处理从低级模式到高级语义的信息，包括曲线和空间位置的编码。该框架简化了 ViT 的计算，便于人类理解，并已证明在消除虚假关联方面具有实用性。 AI

影响为理解和潜在地提高复杂视觉模型的可解释性提供了一个新框架。

排序理由关于解释人工智能模型新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 4 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 关于解释人工智能模型新方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Same-day; Cluster formed today. Ranking reflects the current source set at time of score.

完整方法见我们的编辑标准。

arXiv cs.CV TIER_1 English(EN) · Jinyeong Kim, Junhyeok Kim, Yumin Shim, Joohyeok Kim, Sunyoung Jung, Seong Jae Hwang · 2026-10-23 04:00

Interpreting vision transformers via residual replacement model

arXiv:2509.17401v2 Announce Type: replace Abstract: How do vision transformers (ViTs) represent and process the world? This paper addresses this long-standing question through the first systematic analysis of 6.6K features across all layers, extracted via sparse autoencoders, and…