UniVR: Thinking in Visual Space for Unified Visual Reasoning

arXiv 2026 24.6 method, benchmark, system

TLDR

UniVR uses RL to learn visual reasoning, physical dynamics, and planning from pure visual demonstrations, achieving 25% improvement on a new benchmark.

Reasoning

Strengths include novel integration of reasoning and dynamics from visual data, a large benchmark, and open-source release. Weaknesses are limited methodological details, lack of baseline comparisons, and unverified 'first' claim.

Read-first score

Read-first score 24.6, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 16.

Recency 6%

100

Uses a gentle age decay so recent papers surface without erasing older foundations. 2026

Reproducibility 18%

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=code

Methodology quality 18%

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=benchmark

Topical relevance 29%

22.9

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Citation impact 18%

Uses OpenAlex-shaped citation metadata as a bibliometric attention signal, separate from paper quality. cited_by_count=0

Citation velocity 12%

Citation velocity estimates citations per publication-year to reduce old-paper bias. velocity=0.00

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 48.

Keyword Scores

world model

world dynamics prediction

video world model

model-based reinforcement learning world model

world simulator

generative world model

interactive world model

Resources

Paper