RIVER Bench
RIVER: A Real-Time Interaction Benchmark for Video LLMs
Yansong Shi, Qingsong Zhao, Tianxiang Jiang, Xiangyu Zeng, Yi Wang, Limin Wang · 2026-03-04
- Modality
- Video & streaming
- Data
- Hybrid (real + synthetic/simulated)
- Task · ground truth
- QA/Act. · Ans.+Time+Act.
- Scope
- Benchmark with memory tracks
What it tests
- CECross-modal evidenceDirectly evaluated
- LELong-range evidenceDirectly evaluated
- SCState changesDirectly evaluated
- CFConflicting memoriesNot evaluated
- UQUnanswerable queriesNot evaluated
- MAMemory-grounded actionDirectly evaluated
- TLTemporal localizationDirectly evaluated
- TRTemporal relationsDirectly evaluated
- MEMulti-event integrationDirectly evaluated
Survey notes
Retro-Memory questions target previously observed events with temporal gaps.