Beyond Visual Memory: Mechanistic Diagnostics of Latent Visual Reasoning

arXiv:2606.01287 · cs.CV, cs.AI · Submitted 2026-05-31 · Read on arXiv

cs.CV, cs.AI

Submitted: 2026-05-31

Updated: 2026-08-26

Terminology

• Accuracy — 92× in this paper · explained in 1 episode
• Benchmark — 33× in this paper · explained in 2 episodes
• Alignment — 18× in this paper · explained in 1 episode
• Latent Visual Reasoning — 17× in this paper · explained in 2 episodes
• Embeddings — 10× in this paper · explained in 2 episodes
• GRPO — 8× in this paper · explained in 6 episodes
• Decode — 5× in this paper · explained in 2 episodes
• Latent Reasoning — 4× in this paper · explained in 2 episodes
• Transformer — 4× in this paper · explained in 2 episodes
• Attention Entropy — 3× in this paper · explained in 2 episodes
• Distillation — 3× in this paper · explained in 2 episodes
• Ablation — 2× in this paper · explained in 1 episode
• Core — 2× in this paper · explained in 2 episodes
• Cosine Similarity — 2× in this paper · explained in 3 episodes
• Filler Tokens — 2× in this paper · explained in 2 episodes
• Multimodal Reasoning — 2× in this paper · explained in 5 episodes
• Reinforcement Learning — 2× in this paper · explained in 4 episodes
• Trace — 2× in this paper · explained in 2 episodes
• Ablation Studies — 1× in this paper · explained in 1 episode
• Anchor Tokens — 1× in this paper · explained in 1 episode
• Block Sensitivity Ratio — 1× in this paper
• Chain-of-Thought Prompting — 1× in this paper · explained in 2 episodes
• Consistency — 1× in this paper · explained in 3 episodes
• Latent Space — 1× in this paper · explained in 7 episodes
• Mesoscale Visual Hypotheses — 1× in this paper
• Redundancy — 1× in this paper · explained in 2 episodes
• Role Separation Ratio — 1× in this paper
• Training-Free Framework — 1× in this paper · explained in 2 episodes

Sources

• Qwen3-VL Technical Report
• Qwen2.5-VL Technical Report
• CoLVR: Enhancing Exploratory Latent Visual Reasoning via Contrastive Optimization
• Interleaved Latent Visual Reasoning with Selective Perceptual Modeling
• VLMEvalKit: An Open-Source Toolkit for Evaluating Large Multi-Modality Models
• Training Large Language Models to Reason in a Continuous Latent Space
• Chain-of-Visual-Thought: Teaching VLMs to See and Think Better with Continuous Visual Tokens
• Forest Before Trees: Latent Superposition for Efficient Visual Reasoning
• Proximal Policy Optimization Algorithms
• DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
• Latent Sketchpad: Sketching Visual Thoughts to Elicit Multimodal Reasoning in MLLMs
• Visual Latents Know More Than They Say: Unsilencing Latent Reasoning in MLLMs
• What's Holding Back Latent Visual Reasoning?
• Do Latent Tokens Think? A Causal and Adversarial Analysis of Chain-of-Continuous-Thought
• A Survey on Latent Reasoning

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