EpiCon: Collective Agent Learning through Co-Evolving Multimodal Memory

arXiv:2609.37923 · cs.CV · Submitted 2026-09-29 · Read on arXiv

cs.CV

Submitted: 2026-09-29

Updated: 2026-09-29

Code: https://github.com/deepseek-ai/deepseek-harness

Project page: https://zzzmyyzeng.github.io/EpiCon

Terminology

• Benchmark — 48× in this paper · explained in 2 episodes
• Memory Controller — 9× in this paper
• Ablation — 7× in this paper · explained in 1 episode
• Multi-Agent System — 7× in this paper · explained in 5 episodes
• Agent Memory — 5× in this paper · explained in 1 episode
• Trace — 5× in this paper · explained in 2 episodes
• Self-Evolving Agents — 4× in this paper · explained in 2 episodes
• AI Agents — 2× in this paper · explained in 2 episodes
• Distillation — 2× in this paper · explained in 2 episodes
• Multi-agent system (MAS) — 2× in this paper · explained in 2 episodes
• Multimodal Reasoning — 2× in this paper · explained in 5 episodes
• Provenance — 2× in this paper · explained in 3 episodes
• Winter Conference on Applications of Computer Vision — 2× in this paper
• Abstraction — 1× in this paper · explained in 1 episode
• Adam — 1× in this paper · explained in 2 episodes
• CVF International Conference on Computer Vision — 1× in this paper
• Fidelity — 1× in this paper · explained in 2 episodes
• Fine-Tuning — 1× in this paper · explained in 5 episodes
• Generalization — 1× in this paper · explained in 13 episodes
• Grounding — 1× in this paper · explained in 3 episodes
• Instruction Tuning — 1× in this paper · explained in 2 episodes
• Iterative Refinement — 1× in this paper · explained in 3 episodes
• Knowledge Graph — 1× in this paper · explained in 7 episodes
• Knowledge graphs — 1× in this paper · explained in 2 episodes
• Large Language Model (LLM) — 1× in this paper · explained in 3 episodes
• Multi-agent Collaboration — 1× in this paper · explained in 2 episodes
• Multimodal Dataset — 1× in this paper · explained in 2 episodes
• Reinforcement Learning — 1× in this paper · explained in 4 episodes
• Transformer — 1× in this paper · explained in 2 episodes

Sources

• BabyVision: Visual Reasoning Beyond Language
• Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory
• Improving Factuality and Reasoning in Language Models through Multiagent Debate
• Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks
• MMCOMPOSITION: Revisiting the Compositionality of Pre-trained Vision-Language Models
• MMIG-Bench: Towards Comprehensive and Explainable Evaluation of Multi-Modal Image Generation Models
• ChartNet: A Million-Scale, High-Quality Multimodal Dataset for Robust Chart Understanding
• LowPowAR: Power-Constrained Tone Mapping for Augmented Reality
• Optimizing the Interface Between Knowledge Graphs and LLMs for Complex Reasoning
• MemGPT: Towards LLMs as Operating Systems
• Vision2Code: A Multi-Domain Benchmark for Evaluating Image-to-Code Generation
• Agent KB: Leveraging Cross-Domain Experience for Agentic Problem Solving
• Hierarchical multimodal transformers for Multi-Page DocVQA
• Voyager: An Open-Ended Embodied Agent with Large Language Models
• Agent Workflow Memory
• AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
• OpenSkill: Open-World Self-Evolution for LLM Agents
• WorldBench: A Challenging and Visually Diverse Multimodal Reasoning Benchmark
• Aurora: Unified Video Editing with a Tool-Using Agent
• MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents

Related papers

• Loss Knows Best: Detecting Annotation Errors in Videos via Loss Trajectories
• AnchorWeave: World-Consistent Video Generation with Retrieved Local Spatial Memories
• Benchmarking the Robustness of Foundation Models for Mammography under Domain Shift
• MambaX-Net: Dual-Input Mamba-Enhanced Cross-Attention Network for Longitudinal MRI Segmentation
• TeleOCR: Navigating Document Parsing Across Digital and Camera-Captured Documents
• A Survey on Efficient Vision-Language-Action Models