most citedDEAL: Disentangle and Localize Concept-level Explanations for VLMs

2 citations · 2 across the 5 of their papers we have counts for

collaborators

6 papers

cs.CV2026

Medical AI Encodes a "Feeling of Error": Verifying Cancer Segmentation via Internal Concepts

Mengmeng Ma, Yunxiang Peng, Tang Li +4

Cancer segmentation models can fail silently, generating plausible but incorrect masks that risk missed findings or unnecessary biopsies. A critical question arises: Do AI models "…

cs.CV2026

Inside the Visual Mind: Neuroscience-Motivated Concept Circuits for Interpreting and Steering Vision Transformers

Tang Li, Yanlin Chen, Mengmeng Ma +1

Despite high accuracy, Vision Transformer (ViT) predictions can be driven by spurious cues, raising the need to understand their inner workings before safe deployment. Sparse autoe…

cs.CV2024

SeafloorAI: A Large-scale Vision-Language Dataset for Seafloor Geological Survey

Kien X. Nguyen, Fengchun Qiao, Arthur Trembanis +1

A major obstacle to the advancements of machine learning models in marine science, particularly in sonar imagery analysis, is the scarcity of AI-ready datasets. While there have be…

cs.CV20242 cited

DEAL: Disentangle and Localize Concept-level Explanations for VLMs

Tang Li, Mengmeng Ma, Xi Peng

Large pre-trained Vision-Language Models (VLMs) have become ubiquitous foundational components of other models and downstream tasks. Although powerful, our empirical results reveal…

cs.LG2024

Beyond the Federation: Topology-aware Federated Learning for Generalization to Unseen Clients

Mengmeng Ma, Tang Li, Xi Peng

Federated Learning is widely employed to tackle distributed sensitive data. Existing methods primarily focus on addressing in-federation data heterogeneity. However, we observed th…

cs.LG2024

Adaptive Cascading Network for Continual Test-Time Adaptation

Kien X. Nguyen, Fengchun Qiao, Xi Peng

We study the problem of continual test-time adaption where the goal is to adapt a source pre-trained model to a sequence of unlabelled target domains at test time. Existing methods…