Overview

Latent visual reasoning explores whether models can preserve and manipulate visual evidence in compact internal states instead of spelling out every step in text or repeatedly processing images. Common approaches add continuous visual tokens or latent workspaces, train them with visual-region supervision, answer-based rewards, or distillation, and use predictive world models to anticipate future scenes. These methods are tested on fine-grained perception and spatial benchmarks such as V* and BLINK, as well as video understanding and embodied planning tasks, often aiming to reduce inference time and token use. A central open question is whether the latent states causally contribute grounded evidence to answers, since accuracy gains can also arise from formatting, attention changes, or auxiliary training; interventions and evidence-grounding remain important evaluation tools.

Latest papers

CardsList

95 papers7 papers in the last four weeks, down 46% on the four weeks before. 0.1% of all new papers.

  1. LaST-VLA: Thinking in Latent Spatio-Temporal Space for Vision-Language-Action in Autonomous Driving

    Mar 2, 2026Yuechen Luo, Fang Li, Shaoqing Xu +10VLMs for Autonomous DrivingVision-Language-Action Models

  2. Imagination Helps Visual Reasoning, But Not Yet in Latent Space

    Feb 26, 2026You Li, Chi Chen, Yanghao Li +4Causal Mediation AnalysisVisual Reasoning

  3. CrystaL: Spontaneous Emergence of Visual Latents in MLLMs

    Feb 24, 2026Yang Zhang, Danyang Li, Yuxuan Li +4Vision-Language ModelsLarge Vision-Language Models

  4. Multimodal Latent Reasoning via Hierarchical Visual Cues Injection

    Feb 5, 2026Yiming Zhang, Qiangyu Yan, Borui Jiang +1VLM ReasoningMultimodal Large Language Models

  5. Neurosymbolic Object-Centric Learning with Distant Supervision

    Jun 19, 2025Stefano Colamonaco, David Debot, Giuseppe MarraNeurosymbolic LearningObject-Centric Representation Learning