mlsanityData to 7 Oct 2026

COSMIR: Chain Orchestrated Structured Memory for Iterative Reasoning over Long Context

arXiv 2510.04568 ↗First posted 6 Oct 2025cs.AI cs.LGcited by 1

On the map: LLM reasoning → Reasoning for agents, planning and code

Abstract (arXiv)

Reasoning over very long inputs remains difficult for large language models (LLMs). Common workarounds either shrink the input via retrieval (risking missed evidence), enlarge the context window (straining selectivity), or stage multiple agents to read in pieces. In staged pipelines (e.g., Chain of Agents, CoA), free-form summaries passed between agents can discard crucial details and amplify early mistakes. We introduce COSMIR (Chain Orchestrated Structured Memory for Iterative Reasoning), a chain-style framework that replaces ad hoc messages with a structured memory. A Planner agent first turns a user query into concrete, checkable sub-questions. worker agents process chunks via a fixed micro-cycle: Extract, Infer, Refine, writing all updates to the shared memory. A Manager agent then Synthesizes the final answer directly from the memory. This preserves step-wise read-then-reason benefits while changing both the communication medium (structured memory) and the worker procedure (fixed micro-cycle), yielding higher faithfulness, better long-range aggregation, and auditability. On long-context QA from the HELMET suite, COSMIR reduces propagation-stage information loss and improves accuracy over a CoA baseline.

Key facts. Cited by 1 paper in the map (data through 7 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 35 of its 42 references that are in the map. First posted on arXiv 6 Oct 2025.

Builds on

35 of 42 references.

Closest ancestors · 3

Same direction, newest first

  1. 2025MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent3 Jul
  2. 2023Towards Reasoning in Large Language Models via Multi-Agent Peer Review Collaboration14 Nov
  3. 2023ReConcile: Round-Table Conference Improves Reasoning via Consensus among Diverse LLMs22 Sep

Other references · 32

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  3. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  4. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,155
  5. 2022Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language modelscited by 1,970
  6. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,906
  7. 2023Improving Factuality and Reasoning in Language Models through Multiagent Debatecited by 1,739
  8. 2024Large Language Model based Multi-Agents: A Survey of Progress and Challengescited by 752
  9. 2023YaRN: Efficient Context Window Extension of Large Language Modelscited by 584
  10. 2023MemoryBank: Enhancing Large Language Models with Long-Term Memorycited by 539
  11. 2024∞Bench: Extending Long Context Evaluation Beyond 100K Tokenscited by 360
  12. 2023MedAgents: Large Language Models as Collaborators for Zero-shot Medical Reasoningcited by 311
20 more
  1. 2025From RAG to Memory: Non-Parametric Continual Learning for Large Language Modelscited by 223
  2. 2024Chain of Agents: Large Language Models Collaborating on Long-Context Taskscited by 201
  3. 2022Recurrent Memory Transformercited by 197
  4. 2024InfLLM: Training-Free Long-Context Extrapolation for LLMs with an Efficient Context Memorycited by 192
  5. 2023MAC-SQL: A Multi-Agent Collaborative Framework for Text-to-SQLcited by 158
  6. 2023Walking Down the Memory Maze: Beyond Context Limit through Interactive Readingcited by 129
  7. 2024Self-Discover: Large Language Models Self-Compose Reasoning Structurescited by 108
  8. 2023Examining Inter-Consistency of Large Language Models Collaboration: An In-depth Analysis via Debatecited by 94
  9. 2024MEMORYLLM: Towards Self-Updatable Large Language Modelscited by 89
  10. 2024Beyond the Limits: A Survey of Techniques to Extend the Context Length in Large Language Modelscited by 71
  11. 2024Human-inspired Episodic Memory for Infinite Context LLMscited by 62
  12. 2024Larimar: Large Language Models with Episodic Memory Controlcited by 53
  13. 2025InftyThink: Breaking the Length Limits of Long-Context Reasoning in Large Language Modelscited by 53
  14. 2023Multi-Agent Consensus Seeking via Large Language Modelscited by 50
  15. 2024MASAI: Modular Architecture for Software-engineering AI Agentscited by 49
  16. 2024Disentangling Memory and Reasoning Ability in Large Language Modelscited by 43
  17. 2024Memory Layers at Scalecited by 39
  18. 2024ALR2: A Retrieve-then-Reason Framework for Long-context Question Answeringcited by 25
  19. 2025Sculptor: Empowering LLMs with Cognitive Agency via Active Context Managementcited by 17
  20. 2025Long Context Scaling: Divide and Conquer via Multi-Agent Question-driven Collaborationcited by 7

Built on by

Cited by year

Citing papers per year: 2025: 1.

1 paper.

  1. 2025AI Meets Brain: Memory Systems from Cognitive Neuroscience to Autonomous Agentscited by 12

Similar, not cited

Close in wording, no citation either way

  1. 2026Chow-Liu Ordering for Long-Context Reasoning in Chain-of-Agents
  2. 2025MemR3: Memory Retrieval via Reflective Reasoning for LLM Agents
  3. 2025Look Back to Reason Forward: Revisitable Memory for Long-Context LLM Agents
  4. 2026ByteRover: Agent-Native Memory Through LLM-Curated Hierarchical Context
  5. 2026MemSearch-o1: Empowering Large Language Models with Reasoning-Aligned Memory Growth in Agentic Search
  6. 2026MemFlow: Intent-Driven Memory Orchestration for Small Language Model Agents
  7. 2026StructMem: Structured Memory for Long-Horizon Behavior in LLMs
  8. 2025CogMem: A Cognitive Memory Architecture for Sustained Multi-Turn Reasoning in Large Language Models