AI Paper Atlas

Research map / Video / Video-language understanding

Towards Spatial Supersensing in Video

Shusheng Yang, Jihan Yang, Pinzhi Huang, Ellis L Brown II, Zihao Yang, Yue Yu, Shengbang Tong, Zihan Zheng, Yifan Xu, Muhan Wang, Rob Fergus, Yann LeCun and 2 more

ICLR 2026 · Video › Video-language understanding · approach: Event and anomaly reasoning in video

Read the paper

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train on full VSI-590K mixture instead of the LLaVA-Video-178K subset baseline (ablation)improved Cambrian-1 MLLM · VSI-Benchaverage accuracy: 28.5 -> 63.2Improved
add annotated real-video source ScanNet to the baseline data (ablation)improved Cambrian-1 MLLM · VSI-Benchaverage accuracy: 28.5 -> 56.3Improved
add pseudo-annotated YouTube room-tour images to the baseline data (ablation)improved Cambrian-1 MLLM · VSI-Benchaverage accuracy: 28.5 -> 32.5Improved
add configuration question group to VSI-590K training (ablation)Cambrian-S · VSI-Benchaverage accuracy: 51.9 -> 63.2Improved
add measurement question group to VSI-590K training (ablation)Cambrian-S · VSI-Benchaverage accuracy: 49.7 -> 63.2Improved
add spatiotemporal question group to VSI-590K training (ablation)Cambrian-S · VSI-Benchaverage accuracy: 58.1 -> 63.2Improved
switch from uniform 128-frame sampling to 1 FPS sampling on VSRCambrian-S-7B · VSI-SUPER Recallaccuracy: 26.7 -> 38.3 at 10 min; 23.3 -> 6.0 at 60 minMixed
scale the LLM from 0.5B to 7B in Cambrian-SCambrian-S · VSI-Benchaccuracy: 50.6 -> 67.5Improved
add surprise-driven memory (latent-frame-prediction error) to Cambrian-S for VSR (ablation)Cambrian-S · VSI-SUPER RecallaccuracyImproved
use LFP prediction error as surprise instead of adjacent-frame feature difference (VSR) (ablation)Cambrian-S memory system · VSI-SUPER RecallaccuracyImproved
use LFP prediction error as surprise instead of adjacent-frame feature difference (VSC) (ablation)Cambrian-S event segmentation · VSI-SUPER CountMRAImproved
add surprise-driven event segmentation to Cambrian-S for VSCCambrian-S · VSI-SUPER CountMRA: 38% MRA at 10 min, about 28% at 120 minImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy