Multimodal Agent MemorySubmit a paper

Video Memory

84 systems and 23 benchmarks whose primary modality is video & streaming.

ImageAudio & speechGUI & computer useEmbodied & 3D

How they represent memory

How T2 systems change it

Representation × management tier

T1 Static Memory 27 · 32% T2 Dynamic Memory 54 · 64% T3 Self-Evolving Memory 3 · 4% Total
Source- and modality-specific 12 12 – 24
Cross-represented explicit 3 5 2 10
Internalized latent & parametric 1 7 – 8
Hybrid 11 30 1 42

Each system counted once, at its highest tier. Circle area shows the value. Hover to compare with chance; click to list the systems.

What its benchmarks test

Directly evaluated, of 23 benchmarks in the survey's Table 2
  • Cross-modal evidenceCE13
  • Long-range evidenceLE23
  • State changesSC5
  • Conflicting memoriesCF0
  • Unanswerable queriesUQ0
  • Memory-grounded actionMA6
  • Temporal localizationTL20
  • Temporal relationsTR20
  • Multi-event integrationME23

Benchmarks

All 84 systems