{"ID":6267210,"CreatedAt":"2026-07-10T01:11:38.759438437Z","UpdatedAt":"2026-07-13T01:02:08.706470581Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.08497","arxiv_id":"2607.08497","title":"Cognitive-structured Multimodal Agent for Multimodal Understanding, Generation, and Editing","abstract":"Recent unified multimodal models show a single architecture can jointly perform vision/language understanding and image generation/editing. However, they repeatedly feed all historical visual and textual inputs into a shared context window, limiting long-horizon multimodal dialogue due to visual token explosion and unreliable cross-turn referencing. We propose a Cognitive-structured Multimodal Agent that externalizes visual information into an Episodic Visual Memory and selectively reactivates relevant episodes during reasoning. The agent consists of a Perceptual Abstraction Engine for structured visual abstraction, a Cognitive Retrieval Engine for cross-turn memory retrieval, and a Multimodal Executive Controller for autonomous task inference and action planning. To address the lack of turn-level retrieval supervision in existing datasets, we develop a Unified Scenario Engine that programmatically generates structured multi-turn conversations with fine-grained retrieval annotations, enabling reinforcement learning to optimize abstraction and retrieval policies. We also construct a long-horizon visual-dialogue benchmark stratified by difficulty to evaluate episodic visual recall. Our 8B agent achieves 91.4% retrieval accuracy over 20-turn sessions, surpassing 32B baselines by +8.2% while nearly halving per-turn inference time (23.1s -\u003e 12.7s). We further present the Cognitive-structured Multimodal Agent Harness (CMA-Harness), a tool-augmented deployment of the same cognitive structure integrating persistent multimodal memory, web access, image generation/editing/composition tools, and OpenAI-compatible serving. Structured memory and modular decision-making offer a more scalable, efficient paradigm for long-horizon multimodal agents than monolithic parameter scaling. Code: https://github.com/caseclose/cma-harness ; Project page: https://caseclose.github.io/cma-harness/","short_abstract":"Recent unified multimodal models show a single architecture can jointly perform vision/language understanding and image generation/editing. However, they repeatedly feed all historical visual and textual inputs into a shared context window, limiting long-horizon multimodal dialogue due to visual token explosion and unr...","url_abs":"https://arxiv.org/abs/2607.08497","url_pdf":"https://arxiv.org/pdf/2607.08497v1","authors":"[\"Feng Wang\",\"Canmiao Fu\",\"Zhipeng Huang\",\"Chen Li\",\"Jing Lyu\",\"Ge Li\"]","published":"2026-07-09T13:55:55Z","proceeding":"cs.CV","tasks":"[\"cs.CV\",\"cs.AI\",\"cs.CL\",\"cs.LG\"]","methods":"[\"Reinforcement Learning\"]","has_code":false,"code_links":[{"ID":614083,"CreatedAt":"2026-07-10T01:11:38.759438437Z","UpdatedAt":"2026-07-10T01:11:38.759438437Z","DeletedAt":null,"paper_id":6267210,"paper_url":"https://arxiv.org/abs/2607.08497","paper_title":"Cognitive-structured Multimodal Agent for Multimodal Understanding, Generation, and Editing","repo_url":"https://github.com/caseclose/cma-harness","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}