paper-with-me

홈 › Papers

What Do VLMs NOTICE? A Mechanistic Interpretability Pipeline for Gaussian-Noise-free Text-Image Corruption and Evaluation

2024-06-24 · Michal Golovanevsky, William Rudman, Vedant Palit, Ritambhara Singh, Carsten Eickhoff

Vision-Language Models (VLMs) have gained community-spanning prominence due to their ability to integrate visual and textual inputs to perform complex tasks. Despite their success, the internal decision-making processes of these models remain opaque, posing challenges in high-stakes applications. To address this, we introduce NOTICE, the first Noise-free Text-Image Corruption and Evaluation pipeline for mechanistic interpretability in VLMs. NOTICE incorporates a Semantic Minimal Pairs (SMP) framework for image corruption and Symmetric Token Replacement (STR) for text. This approach enables semantically meaningful causal mediation analysis for both modalities, providing a robust method for analyzing multimodal integration within models like BLIP. Our experiments on the SVO-Probes, MIT-States, and Facial Expression Recognition datasets reveal crucial insights into VLM decision-making, identifying the significant role of middle-layer cross-attention heads. Further, we uncover a set of ``universal cross-attention heads'' that consistently contribute across tasks and modalities, each performing distinct functions such as implicit image segmentation, object inhibition, and outlier inhibition. This work paves the way for more transparent and interpretable multimodal systems.

📄 PDF Abstract BibTeX arXiv:2406.16320

Code (1)

wrudman/NOTICE 공식 구현 pytorch

Tasks

Decision MakingFacial Expression RecognitionImage SegmentationSemantic Segmentation

Methods 이 논문이 사용한 방법론

SET Dynamic Sparse Training method where weight mask is updated randomly periodically
BLIP Vision-Language Pre-training (VLP) has advanced the performance for many vision-language tasks. However, most existing pre-trained models only excel in either understanding-based…

Similar Papers 제목 키워드 기반

Do VLMs Have Bad Eyes? Diagnosing Compositional Failures via Mechanistic Interpretability

2025-08-20 · Ashwath Vaithinathan Aravindan, Abha Jha, Mihir Kulkarni arxiv

Vision-Language Models (VLMs) have shown remarkable performance in integrating visual and textual information for tasks such as image captioning and visual question answering. However, these models struggle with composit…

Visual Question AnsweringImage Captioning

Why Is Spatial Reasoning Hard for VLMs? An Attention Mechanism Perspective on Focus Areas

2025-03-03 · Shiqi Chen, Tongyao Zhu, Ruochen Zhou, Jinghan Zhang 외

Large Vision Language Models (VLMs) have long struggled with spatial reasoning tasks. Surprisingly, even simple spatial reasoning tasks, such as recognizing "under" or "behind" relationships between only two objects, pos…

Spatial Reasoning

Counting Circuits: Mechanistic Interpretability of Visual Reasoning in Large Vision-Language Models

2026-03-19 · Liwei Che, Zhiyu Xue, Yihao Quan, Benlin Liu 외 arxiv

Counting serves as a simple but powerful test of a Large Vision-Language Model's (LVLM's) reasoning; it forces the model to identify each individual object and then add them all up. In this study, we investigate how LVLM…

Visual Reasoning

Interpretable and Steerable Concept Bottleneck Sparse Autoencoders

2025-12-11 · Akshay Kulkarni, Tsui-Wei Weng, Vivek Narayanaswamy, Shusen Liu 외 arxiv

Sparse autoencoders (SAEs) promise a unified approach for mechanistic interpretability, concept discovery, and model steering in LLMs and LVLMs. However, realizing this potential requires learned features to be both inte…

Image Generation

Mechanistic Finetuning of Vision-Language-Action Models via Few-Shot Demonstrations

2025-11-27 · Chancharik Mitra, Yusen Luo, Raj Saravanan, Dantong Niu 외 arxiv

Vision-Language Action (VLAs) models promise to extend the remarkable success of vision-language models (VLMs) to robotics. Yet, unlike VLMs in the vision-language domain, VLAs for robotics require finetuning to contend …