PROBE: Diagnosing Residual Concept Capacity in Erased Text-to-Video Diffusion Models

• URL: http://arxiv.org/abs/2603.21547v1
• Date: Mon, 23 Mar 2026 04:01:13 GMT
• Title: PROBE: Diagnosing Residual Concept Capacity in Erased Text-to-Video Diffusion Models
• Abstract summary: Concept techniques for text-to-video (T2V) diffusion models report substantial suppression of sensitive content.<n>We introduce PROBE, a diagnostic protocol that quantifies the textitre potential erased concepts in T2V models.
• Score: 9.957850061310054
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Concept erasure techniques for text-to-video (T2V) diffusion models report substantial suppression of sensitive content, yet current evaluation is limited to checking whether the target concept is absent from generated frames, treating output-level suppression as evidence of representational removal. We introduce PROBE, a diagnostic protocol that quantifies the \textit{reactivation potential} of erased concepts in T2V models. With all model parameters frozen, PROBE optimizes a lightweight pseudo-token embedding through a denoising reconstruction objective combined with a novel latent alignment constraint that anchors recovery to the spatiotemporal structure of the original concept. We make three contributions: (1) a multi-level evaluation framework spanning classifier-based detection, semantic similarity, temporal reactivation analysis, and human validation; (2) systematic experiments across three T2V architectures, three concept categories, and three erasure strategies revealing that all tested methods leave measurable residual capacity whose robustness correlates with intervention depth; and (3) the identification of temporal re-emergence, a video-specific failure mode where suppressed concepts progressively resurface across frames, invisible to frame-level metrics. These findings suggest that current erasure methods achieve output-level suppression rather than representational removal. We release our protocol to support reproducible safety auditing. Our code is available at https://github.com/YiweiXie/PRObingBasedEvaluation.

Related papers

• Interpreting Object-Dependent Concept Brittleness in Text-to-Image Diffusion Models [105.53190946111586]
  We term this phenomenon object-dependent concept brittleness.<n>We present an interpretability-oriented framework to audit and minimally correct these failures.<n>We evaluate the proposed framework on style and attribute failure cases across multiple diffusion backbones.
  arXiv  Detail & Related papers  (2026-09-09T09:07:40Z)
• EraseSAE: Surgical Concept Erasure in Text-to-Video Diffusion Models via Sparse Autoencoders [103.46709005379591]
  Concept erasure offers a principled remedy by removing unwanted semantics from pretrained models.<n>Existing approaches typically operate at a coarse quality, misaligned with the finegrained, distributed nature of concept representations.<n>We propose EraseSAE, a novel framework that leverages sparse autoencoders to achieve surgical concept erasure.
  arXiv  Detail & Related papers  (2026-09-03T10:23:37Z)
• DriftScope: Measuring The Hidden Effects of Diffusion Model Adaptation [50.584834483403675]
  Adapting pre-trained text-to-image diffusion models, whether to learn new visual concepts or erase unwanted ones, is routinely evaluated on its intended effects alone.<n>We demonstrate that adaptation systematically damages semantically unrelated concepts in ways that aggregate metrics structurally cannot surface.<n>We introduce DriftScope, a prompt-level diagnostic tool that takes any two model checkpoints and returns a ranked list of tokens whose visual concepts have shifted most between them.
  arXiv  Detail & Related papers  (2026-06-30T20:57:51Z)
• Speculative Refinement: A Hybrid Autoregressive Diffusion Decoding Strategy and Its Behavior Across Benchmarks [1.9800951131982487]
  We use Speculative Refinement (SpecRef) to warm-start a masked diffusion language model from an AR draft.<n>We evaluate SpecRef across six benchmarks (HumanEval, MBPP, GSM8K, BBH, ARC-Challenge, HellaSwag) with three distinct evaluation protocols.<n>These observations apply to any multi-stage or non-autoregressive generation pipeline and point toward more diagnostic evaluation practices.
  arXiv  Detail & Related papers  (2026-06-25T18:52:24Z)
• Beyond Reconstruction: Reconstruction-to-Vector Diffusion for Hyperspectral Anomaly Detection [9.078979356836193]
  Hyperspectral Anomaly Detection (HAD) excels at identifying sparse targets in complex scenes.<n>Existing models remain trapped in a scalar "reconstruction-as-endpoint" paradigm.<n>We propose Reconstruction-to-Diffusion (R2VD), which redefines reconstruction as a purification origin.<n>R2VD establishes a new state-of-the-art, delivering exceptional target detectability background.
  arXiv  Detail & Related papers  (2026-04-13T12:31:50Z)
• Interpreting Video Representations with Spatio-Temporal Sparse Autoencoders [3.3192479135000426]
  We present the first systematic study of Sparseencoencoders (SAEs) on video representations.<n>Standard SAEs decompose into interpretable, monosemantic features but destroy temporal coherence.<n>TopK selection produces unstable feature assignments across frames, reducing autocorrelation by 36%.
  arXiv  Detail & Related papers  (2026-04-05T01:08:19Z)
• M-ErasureBench: A Comprehensive Multimodal Evaluation Benchmark for Concept Erasure in Diffusion Models [13.885985776728559]
  M-ErasureBench is a novel framework that benchmarks concept erasure methods across three input modalities.<n> IRECE is a plug-and-play module that localizes target concepts via cross-attention and perturbs the associated latents during denoising.
  arXiv  Detail & Related papers  (2025-12-28T10:58:36Z)
• Enhancing Dual Network Based Semi-Supervised Medical Image Segmentation with Uncertainty-Guided Pseudo-Labeling [5.1962665598872135]
  This paper proposes a novel semi-supervised 3D medical image segmentation framework based on a dual-network architecture.<n>Specifically, we investigate a Cross Consistency Enhancement module using both cross pseudo and entropy-filtered supervision to reduce the noisy pseudo-labels.<n>In addition, we use a self-supervised contrastive learning mechanism to align uncertain voxel features with reliable class prototypes.
  arXiv  Detail & Related papers  (2025-09-16T13:40:20Z)
• Erased or Dormant? Rethinking Concept Erasure Through Reversibility [6.895055915600732]
  We evaluate two representative concept erasure methods, Unified Concept Editing and Erased Stable Diffusion.<n>We show that erased concepts often reemerge with substantial visual fidelity after minimal adaptation.<n>Our findings reveal critical limitations in existing concept erasure approaches.
  arXiv  Detail & Related papers  (2025-05-22T03:26:46Z)
• T2VShield: Model-Agnostic Jailbreak Defense for Text-to-Video Models [88.63040835652902]
  Text to video models are vulnerable to jailbreak attacks, where specially crafted prompts bypass safety mechanisms and lead to the generation of harmful or unsafe content.<n>We propose T2VShield, a comprehensive and model agnostic defense framework designed to protect text to video models from jailbreak threats.<n>Our method systematically analyzes the input, model, and output stages to identify the limitations of existing defenses.
  arXiv  Detail & Related papers  (2025-04-22T01:18:42Z)
• Reliable and Efficient Concept Erasure of Text-to-Image Diffusion Models [76.39651111467832]
  We introduce Reliable and Efficient Concept Erasure (RECE), a novel approach that modifies the model in 3 seconds without necessitating additional fine-tuning. To mitigate inappropriate content potentially represented by derived embeddings, RECE aligns them with harmless concepts in cross-attention layers. The derivation and erasure of new representation embeddings are conducted iteratively to achieve a thorough erasure of inappropriate concepts.
  arXiv  Detail & Related papers  (2024-07-17T08:04:28Z)
• Six-CD: Benchmarking Concept Removals for Benign Text-to-image Diffusion Models [58.74606272936636]
  Text-to-image (T2I) diffusion models have shown exceptional capabilities in generating images that closely correspond to textual prompts.<n>The models could be exploited for malicious purposes, such as generating images with violence or nudity, or creating unauthorized portraits of public figures in inappropriate contexts.<n> concept removal methods have been proposed to modify diffusion models to prevent the generation of malicious and unwanted concepts.
  arXiv  Detail & Related papers  (2024-06-21T03:58:44Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.