Skip to content

Suggestion: Add PersonaVLM and Light-Omni (Multimodal Long-tern Personalized Agent Memory) #6

Description

@NickClawZ

Hi!
I'd like to recommend two recent works on multimodal agent memory for the awesome list:
PersonaVLM: Long-Term Personalized Multimodal LLMs (focuses on long-term personalization, CVPR 2026 Highlight) and
Light-Omni: Reflex over Reasoning in Agentic Video Understanding with Long-Term Memory (focuses on efficient reflexive memory for video understanding).

Here are the links:
PersonaVLM: https://github.com/MiG-NJU/PersonaVLM
Light-Omni: https://github.com/Clare-Nie/Light-Omni

Would be great if they could be included in the multimodal/video agent sections. Let me know if you'd prefer I submit a PR directly! Thanks!

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions