PersistBench, a NeurIPS Spotlight benchmark, tests visual memory in camera-controllable 4D video models
weichiuma · x · 2026-10-20
- A NeurIPS 2026 E&D Spotlight (1 of 33) paper introduces PersistBench, a benchmark asking: can 4D foundation models remember?
- It evaluates visual memory in camera-controllable video models (RecamMaster, Gen3C, etc.), 4D reconstruction models, and video-to-360° models — whether they retain object state after it leaves the camera's view.
- The key challenge, obtaining real-world ground truth for out-of-view objects, is sidestepped by a simple construction trick the authors introduce.
- Project page, leaderboard, GitHub repo, and Hugging Face dataset are all publicly available.
More from Research
- MoCaP: zero-shot mobile manipulation for robots from passive human videos — sherryyangML · 2026-10-20
- Foundation model trained on 40M face photos predicts cancer outcomes and biological age — EricTopol · 2026-10-20
- Why GRPO collapses on multi-turn agent training, and how to stabilize advantage estimation — burkov · 2026-10-20
- SJTU's MSAR Residual Architecture Delivers 1.58x Iso-Loss Training Speedup Over Standard Residuals — burny_tech · 2026-10-20
- From memory capacity to transformers: outer-product stores cap out near 2×key dimension — burny_tech · 2026-10-20
- UIUC paper: Fisher-selected 0.1% of parameters recovers 85% of full RL training gains — burny_tech · 2026-10-20