# Reason, Then Re-reason: Cross-view Revisiting Improves Spatial Reasoning

Source: [Reason, Then Re-reason: Cross-view Revisiting Improves Spatial Reasoning](<https://www.sota2.com/research/paper/6a2a4136d33dfb23e783943e>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-06-14T06:00:01.338Z

Authors: Chaofan Ma, Zhenjie Mao, Yuhuan Yang, Fanqin Zeng, Yue Shi, Yingjie Zhou, Xiaofeng Cao, Jiangchao Yao

Published: 2026-06-10T05:52:14.000Z

## Abstract

Spatial reasoning from egocentric videos is inherently challenging because the observable evidence is constrained by the camera trajectory. Existing methods rely on single-turn inference, forcing models to resolve geometric ambiguity through semantic priors rather than verifiable evidence. We argue that spatial reasoning should be revisitable: conclusions formed under limited evidence should remain open to revision when complementary viewpoints become available. Building on this insight, we propose Reason, then Re-reason (ReRe), a training-free, inference-time framework with two phases: in the Reason Phase, an MLLM forms a spatial hypothesis from the original video; in the Re-reason Phase, it verifies or revises the hypothesis by observing a synthesized novel-view video. To enable effective cross-view revisiting, we design a Geometry-to-Video pipeline that renders strategically complementary novel views from predicted 3D geometry. These views feature an elevated, oblique perspective with scene-spanning coverage, while preserving the MLLM's native video interface without architectural modifications. Extensive evaluations on VSI-Bench and STI-Bench demonstrate that ReRe substantially boosts open-source MLLMs to rival proprietary state-of-the-art performance. Project page: https://zhenjiemao.github.io/ReRe/

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Static Spatial Reasoning on STI-Bench Static](<https://www.sota2.com/research/sota/static-spatial-reasoning-on-sti-bench-static.md>): Static Spatial Reasoning; STI-Bench Static; Avg. Score: 34.8.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2606.11683>)

[PDF](<https://arxiv.org/pdf/2606.11683.pdf>)