Towards Explainable Industrial Anomaly Detection via Knowledge-Guided Latent Reasoning

arXiv:2602.09850 · cs.CV · Submitted 2026-02-10 · Read on arXiv

cs.CV

Submitted: 2026-02-10

Updated: 2026-10-21

Code: https://github.com/chenpeng052/Reason-IAD

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Latent Reasoning — 3× in this paper · explained in 3 episodes
• Accuracy — 1× in this paper · explained in 1 episode
• Core — 1× in this paper · explained in 2 episodes
• Interpretability — 1× in this paper · explained in 15 episodes
• Latent Space — 1× in this paper · explained in 7 episodes
• MLLMs — 1× in this paper · explained in 2 episodes

Sources

• MMAD: A Comprehensive Benchmark for Multimodal Large Language Models in Industrial Anomaly Detection
• AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detection
• IAD-R1: Reinforcing Consistent Reasoning in Industrial Anomaly Detection
• AnomalyR1: A GRPO-based End-to-end MLLM for Industrial Anomaly Detection
• AD-FM: Multimodal LLMs for Anomaly Detection via Multi-Stage Reasoning and Fine-Grained Reward Optimization
• Latent Chain-of-Thought for Visual Reasoning
• Latent Visual Reasoning
• VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learning
• Sherlock: Self-Correcting Reasoning in Vision-Language Models
• SRPO: Enhancing Multimodal LLM Reasoning via Reflection-Aware Reinforcement Learning
• Think Silently, Think Fast: Dynamic Latent Compression of LLM Reasoning Chains
• Seek in the Dark: Reasoning via Test-Time Instance-Level Policy Gradient in Latent Space
• Reasoning Beyond Language: A Comprehensive Survey on Latent Chain-of-Thought Reasoning
• Reasoning to Learn from Latent Thoughts
• Scaling up Test-Time Compute with Latent Reasoning: A Recurrent Depth Approach
• Training Large Language Models to Reason in a Continuous Latent Space
• Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens
• ThinkAct: Vision-Language-Action Reasoning via Reinforced Visual Latent Planning
• Soft Thinking: Unlocking the Reasoning Potential of LLMs in Continuous Concept Space
• Reasoning Within the Mind: Dynamic Multimodal Interleaving in Latent Space

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models