跳到正文
研踪PaperTrail
← 研踪 PaperTrail

arXiv:2603.06533 · v2

NEGATE: Constrained Semantic Guidance for Linguistic Negation in Text-to-Video Diffusion

首次提交 2026-03-07 02:21
版本更新 2026-11-18 01:23
雷达观测 2026-11-18 12:34
arXiv ↗PDF ↗

摘要

Negation is a fundamental linguistic operator, yet it remains inadequately modeled in vision-language generation systems. In this work, we present a formal treatment of linguistic negation in vision-language models by modeling it as a structured feasibility constraint on semantic guidance within diffusion dynamics. Rather than introducing heuristics or retraining model parameters, we reinterpret classifier-free guidance as defining a semantic update direction and enforce negation by projecting the update onto a convex constraint set derived from linguistic structure. This formulation provides a unified framework for handling diverse negation phenomena, including object absence, graded non-inversion semantics, multi-negation composition, and scope-sensitive disambiguation. Our approach is training-free, compatible with pretrained diffusion backbones, and defined on the classifier-free guidance update, which lets it operate over temporally evolving video trajectories. In addition, we introduce a structured negation-centric benchmark suite that isolates distinct linguistic failure modes in generative systems, to further research in this area. Experiments demonstrate that our method achieves robust negation compliance while preserving visual fidelity and structural coherence, moving the treatment of linguistic negation in generative vision-language models beyond representation-level evaluation.

作者备注:51 pages, 32 figures, v2: corrected typos and more results added

AI 论文解读

基于摘要 · 非全文审阅

此版本尚未生成 AI 解读。解读由后台任务生成并按论文版本保存,登录后共享已有结果。

版本差异

本站首次收录为 v2,暂无更早版本可对照。

观测到的版本

相邻记录可能跳过中间版本;首次发现 v2 及以上也会标为版本更新。

v2 · 2026-11-18 01:23 · 最新观测版本

NEGATE: Constrained Semantic Guidance for Linguistic Negation in Text-to-Video Diffusion

Negation is a fundamental linguistic operator, yet it remains inadequately modeled in vision-language generation systems. In this work, we present a formal treatment of linguistic negation in vision-language models by modeling it as a structured feasibility constraint on semantic guidance within diffusion dynamics. Rather than introducing heuristics or retraining model parameters, we reinterpret classifier-free guidance as defining a semantic update direction and enforce negation by projecting the update onto a convex constraint set derived from linguistic structure. This formulation provides a unified framework for handling diverse negation phenomena, including object absence, graded non-inversion semantics, multi-negation composition, and scope-sensitive disambiguation. Our approach is training-free, compatible with pretrained diffusion backbones, and defined on the classifier-free guidance update, which lets it operate over temporally evolving video trajectories. In addition, we introduce a structured negation-centric benchmark suite that isolates distinct linguistic failure modes in generative systems, to further research in this area. Experiments demonstrate that our method achieves robust negation compliance while preserving visual fidelity and structural coherence, moving the treatment of linguistic negation in generative vision-language models beyond representation-level evaluation.

查看此版本解读 →查看此版本原文 ↗