Metraipolis › archive › dc_03ca0d97e32e

MemEvo: Automatic Discovery of Streaming Video Memory Mechanisms — arXiv:2609.36581

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.36581
License (as archived)
CC0-1.0
Archived
2026-10-07T16:22:45.310Z
SHA-256 of the archived text
ec50021b6185adc65ec8943bab69f56a0c3aa65616d3f82d6daaf975b48234cf
Stable URL of this page
https://metraipolis.com/documento/dc_03ca0d97e32e

How to cite

MemEvo: Automatic Discovery of Streaming Video Memory Mechanisms — arXiv:2609.36581 — arXiv (en), https://arxiv.org/abs/2609.36581 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_03ca0d97e32e (Metraipolis archive dc_03ca0d97e32e, sha256 ec50021b6185adc65ec8943bab69f56a0c3aa65616d3f82d6daaf975b48234cf)
BibTeX
@misc{dc03ca0d97e32e,
  title = {MemEvo: Automatic Discovery of Streaming Video Memory Mechanisms — arXiv:2609.36581},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.36581},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_03ca0d97e32e; sha256 ec50021b6185adc65ec8943bab69f56a0c3aa65616d3f82d6daaf975b48234cf},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.36581. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Query-agnostic streaming video understanding requires vision-language models to continuously compress an indefinitely growing visual stream into a bounded memory before future queries are known. The performance depends critically on the memory mechanism--what observations to preserve, how to represent and consolidate them, and what information to retrieve when a query eventually arrives. Rather than designing a single memory architecture by hand, we formulate memory design as a search problem over executable memory programs. We introduce a lightweight domain-specific language that expresses memory mechanisms through structured primitives for representation, admission, retention, consolidation, budgeting, and retrieval, while enforcing causal and bounded-memory constraints. Although structured, the derived program space remains large and contains heterogeneous, conditionally dependent design choices whose effects can only be assessed via downstream execution. We therefore propose MemEvo, an LLM-driven auto-research framework that uses pretrained LLM as a semantics-aware proposal model to iteratively generate and refine candidate memory programs based on accumulated experimental feedback. At runtime, a deterministic evaluation pipeline validates and evaluates each candidate, while the underlying vision-language model remains frozen throughout discovery. We finally produce a training-free, bounded-memory mechanism. Extensive experiments on StreamingBench and OVO-Bench demonstrate strong streaming video understanding performance together with substantial context and inference efficiency.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-30 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.36581 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-07T15:54:53.513Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-30?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).