Face recognition templates are compact identity representations, yet they also encode rich semantic information about facial appearance. Prior work has shown that templates can be inverted to images or indirectly manipulated through image-editing pipelines, but direct semantic editing in template space remains largely unexplored. Existing interpretability methods for face recognition often rely on manual neuron inspection or predefined attribute labels, limiting scalability and semantic flexibility. To address this gap, we propose SCOUT (Semantic Concept Discovery for Open-VocabUlary Editing of Face Recognition Templates), an end-to-end framework for discovering and directly manipulating semantic concepts in face recognition templates using mechanistic interpretability. SCOUT learns sparse template representations, generates semantic hypotheses for latent features from natural-language descriptions, and validates their stability. The resulting features act as controllable semantic directions for direct editing, avoiding costly edit--re-encode pipelines. Experiments with face recognition models using CNN, ViT, and Swin backbones show that SCOUT discovers interpretable concepts beyond standard attribute labels and enables controllable, identity-aware template manipulation with negligible impact on identity matching. We further show that edited templates can subsequently be decoded with independent inversion models for visualization and evaluation.
Mechanistic interpretability has recently expanded to Vision Transformers (ViTs), with Sparse Autoencoders (SAEs) increasingly used as post-hoc tools to decompose internal representations into sparse and more interpretable features. However, because post-hoc SAEs are trained on frozen representations after the ViT has already been optimized, their latent features are not directly aligned with the downstream classification objective. We introduce SpIn-ViT, a framework that jointly trains a pretrained ViT and a modified SAE end-to-end, directly aligning sparse patch-level representations with image classification. SpIn-ViT learns semantically coherent neuron activations that localize meaningful image regions while maintaining competitive predictive performance. We evaluate SpIn-ViT across nine image-classification benchmarks using classification accuracy, quantitative interpretability metrics, AI-based and Human evaluations. Compared with the previous state-of-the-art post-hoc SAE method, SpIn-ViT achieves 8.84% higher average classification accuracy, an AI-based interpretability score nearly four times as high, and a human-evaluation score more than twice as high. We further extract interpretable rule-sets using the SAE neurons to create neurosymbolic models which achieve 5.97% higher average classification accuracy while requiring a 58.8\% smaller rule-set than the neurosymbolic models created from the SOTA post-hoc SAE method.
Urban flooding poses an escalating threat to transportation infrastructure, yet no operational system provides real-time, street-level flood-depth estimates at centimeter resolution. This paper presents three vision-language models fine-tuned for continuous flood-depth estimation from street-level imagery: FloodLlama-Dense, a fully fine-tuned QLoRA baseline, and FloodLlama-MI5 and FloodLlama-MI6, interpretability-guided sparse variants that fine-tune only the top five and six causally relevant cross-attention layers identified through mechanistic interpretability analysis, respectively. Training uses an approximately 610,000-image subset of a 2.81-million-image synthetic corpus generated in Unreal Engine 5. The dataset combines single-vehicle subsets with 5 cm depth increments and mixed-vehicle subsets with 1 cm depth increments, spanning seven vehicle types, four weather conditions, and flood depths from 0 to 40 cm. FloodLlama-Dense achieves an MAE of 0.40 cm, an RMSE of 1.97 cm, and an Acc@5cm of 97.59%. Mechanistic interpretability analysis combining linear probing, logit lens, centered kernel alignment (CKA), and cross-attention entropy reveals a two-stage adaptation pattern: layers L13-L22 restructure visual representations, while depth first becomes linearly decodable at layer L23. FloodLlama-MI5 and FloodLlama-MI6 leverage this insight by fine-tuning only five or six of the eight cross-attention layers, achieving an 86-88% reduction in trainable parameters (6.55-7.86 million versus 54.4 million) with minimal accuracy loss. On a real-world benchmark, FloodLlama-MI6 achieves 98.62% accuracy, compared with 86.61% for the published STURM-FloodDepth baseline.
Yiming Tang, Qinglin Qi, Zhaoqian Yao +2cs.CV cs.AI
Mechanistic interpretability has made significant strides in understanding neural network representations, with sparse dictionary learning (SDL) methods, most prominently sparse autoencoders, as a central paradigm. However, recent work has reported several limitations of this paradigm: SDL objectives are non-identifiable; SDL methods rely heavily on the Linear Representation Hypothesis; and a growing body of evidence points to concepts that are encoded non-linearly and are therefore not expressible as any single direction. We hypothesise that a different route to monosemanticity is available. Biological visual systems exhibit highly selective neurons organised into hierarchies of increasing abstraction, and this organisation emerges from local, layer-wise learning rules rather than from a global error signal; we therefore ask whether a biologically plausible learning algorithm will likewise yield monosemantic neurons. To test this, we propose Group-Contrastive Forward-Forward (GCFF), a forward-forward training algorithm that combines class-specific routing with within-class contrastive objectives, reaching monosemanticity through architectural constraints rather than sparsity. Because GCFF attaches multiple non-linear layers to the representation under study, its neurons can therefore capture the non-linear concepts. On CLIP representations, a single trained GCFF module recovers monosemantic neurons whose abstraction increases progressively with depth, reaching environmental properties that hold independently of an image's foreground, without any sparsity constraint or supervision of abstraction level. We further demonstrate that GCFF can train networks from scratch, achieving state-of-the-art performance among forward-forward algorithms on various image classification benchmarks.
Multimodal diffusion transformers (MM-DiTs) have emerged as the prevalent backbone for modern text-to-image generation systems. However, they exhibit critical alignment vulnerabilities, systematically manifesting severe stereotype biases even under benign prompts. This poses a significant risk of algorithmic discrimination in deployed systems. Since most existing mitigation strategies were tailored for legacy U-Net architectures, the precise remediation of these vulnerabilities in MM-DiTs remains a critical open challenge. In this work, we first investigate the root cause of this vulnerability via mechanistic analysis. We reveal that bias representations in MM-DiTs are not uniformly distributed across depth, but are mediated by a sparse set of layers functioning as internal semantic binding hubs. These hubs exhibit a stage-wise propagation driving bias manifestation: early hubs establish the structural templates susceptible to bias, middle hubs actively extract core stereotypical concepts from textual conditioning, and late hubs globally solidify these biases through visual self-attention. Leveraging these architectural insights, we propose FairFlow, an intrinsic, mechanism-guided mitigation framework. FairFlow acts as an internal regulator by employing sparse steering: it learns attribute-specific fair directions and injects them exclusively at the identified semantic hubs within a constrained inference window. Evaluations on FLUX.1-dev and Stable Diffusion~3 demonstrate that FairFlow effectively neutralizes these stereotypical vulnerabilities across gender, race, and intersectional settings, achieving an optimal fairness-fidelity balance. With near-zero inference overhead and robustness to complex prompts, FairFlow provides a lightweight and practical bias mitigation for large-scale deployed MM-DiT systems. Code and datasets will be publicly released upon acceptance.
Bohan Liu, Wenqian Ye, Guangzhi Xiong +3cs.CV cs.CL
Models trained via Contrastive Language-Image Pretraining (CLIP) serve as the foundational vision encoders for most modern Large Vision Language Models (LVLMs). Despite their widespread adoption, CLIP models exhibit a critical yet underexplored failure mode: irrelevant text appearing within images confounds visual representations, biasing them toward lexical meaning rather than true visual semantics. This robustness issue, commonly described as a Typographic Attack (TA), exposes a vulnerability that poses a significant risk to safety-critical applications such as autonomous driving. To achieve interpretable and effective robustness against TA, we propose a novel, training-free mechanistic interpretability method. Our method provides sampling-based interpretations of hidden state representations and quantitatively attributes semantic versus lexical focus to individual attention heads. Through probabilistic analysis and circuit mining, we isolate specific Vision Transformer (ViT) components that disproportionately encode lexical information, thereby identifying the mechanistic source of TA. We further show that simple interventions applied directly to the identified circuits, without any additional training, can substantially improve robustness against Typographic Attacks in object classification. These interventions, such as selective adjustment of attention weights, also outperform both supervised and training-free defense methods. Our experiments demonstrate that applying the proposed intervention to the vision encoders of several state-of-the-art LVLMs yields substantial gains in Visual Question Answering accuracy under Typographic Attack interference on RIO-Bench. These results confirm both the efficacy and the generalizability of our mechanistic approach. Code is released at https://github.com/Liu-524/SamplingTAR.