Inference-Time Concept Suppression and Video-Centric Evaluation for Text-to-Video Models
This paper proposes SIRUS, a training-free inference-time framework for concept-level unlearning in text-to-video (T2V) models. SIRUS localizes target-related prompt evidence and suppresses target expression during sampling without updating the text encoder or denoising network. A video-oriented evaluation framework is introduced to separately measure target forgetting, non-target preservation, video quality, jailbreak robustness, and efficiency. On CogVideoX, SIRUS achieves 70.4% average forgetting success and 25.7% average frame hit, compared to 44.4%/47.2% for VideoEraser, while reducing the average VBench quality drop from -0.043 to -0.016. Transfer experiments on Wan2.2 suggest SIRUS generalizes across modern T2V backbones.
-->
[Submitted on 15 Jul 2026]
Title:Inference-Time Concept Suppression and Video-Centric Evaluation for Text-to-Video Models
View a PDF of the paper titled Inference-Time Concept Suppression and Video-Centric Evaluation for Text-to-Video Models, by Wenxuan Chen and 1 other authors
View PDF HTML (experimental)
Abstract:Text-to-video (T2V) generators can synthesize realistic and temporally coherent videos, but controllably removing a target concept from a generator remains difficult.
Unlike text-to-image concept erasure, T2V unlearning must suppress a target concept that may persist across frames while preserving non-target subjects, actions, scenes, and temporal structure.
We propose \textbf{SIRUS}, a training-free inference-time framework for concept-level T2V unlearning.
Given textual aliases of a target concept, SIRUS localizes target-related prompt evidence and suppresses target expression during sampling, without updating the text encoder or denoising network.
We further introduce a video-oriented evaluation framework for T2V unlearning that separately measures target forgetting, non-target preservation, video quality, jailbreak robustness, and efficiency, using video-level failure criteria, frame-level residue statistics, paired preservation analysis, VBench-based quality diagnostics, and deployment overhead measurement.
Across five safety, object, and style concepts on CogVideoX, SIRUS reaches 70.4\% average forgetting success and 25.7\% average frame hit, compared with 44.4\% / 47.2\% for VideoEraser, while reducing the average VBench quality drop from -0.043 to -0.016, yielding the strongest forgetting-quality trade-off among fully evaluated baselines.
Transfer experiments on Wan2.2 further suggest that SIRUS generalizes across modern T2V backbones.
Comments: 29 pages, 9 figures, 11 tables
Subjects:
Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
Cite as: arXiv:2607.14194 [cs.CV]
(or arXiv:2607.14194v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2607.14194
arXiv-issued DOI via DataCite (pending registration)
Submission history
From: Wenxuan Chen [view email] [v1] Wed, 15 Jul 2026 16:26:09 UTC (12,633 KB)
Full-text links:
Access Paper:
View a PDF of the paper titled Inference-Time Concept Suppression and Video-Centric Evaluation for Text-to-Video Models, by Wenxuan Chen and 1 other authors
View PDF
HTML (experimental)
TeX Source
view license
Current browse context:
cs.CV
new | recent | 2026-07
Change to browse by:
cs cs.LG
References & Citations
NASA ADS
Google Scholar
Semantic Scholar
Loading...
Data provided by:
Bibliographic Tools
Bibliographic and Citation Tools
Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
Connected Papers Toggle
Connected Papers (What is Connected Papers?)
Litmaps Toggle
Litmaps (What is Litmaps?)
scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
Code, Data, Media
Code, Data and Media Associated with this Article
alphaXiv Toggle
alphaXiv (What is alphaXiv?)
Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub Toggle
DagsHub (What is DagsHub?)
GotitPub Toggle
Gotit.pub (What is GotitPub?)
Huggingface Toggle
Hugging Face (What is Huggingface?)
ScienceCast Toggle
ScienceCast (What is ScienceCast?)
Demos
Demos
Replicate Toggle
Replicate (What is Replicate?)
Spaces Toggle
Hugging Face Spaces (What is Spaces?)
Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
Related Papers
Recommenders and Search Tools
Link to Influence Flower
Influence Flower (What are Influence Flowers?)
Core recommender toggle
CORE Recommender (What is CORE?)
Author
Venue
Institution
Topic
About arXivLabs
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)