Sign up to save your podcasts
Or
Share Memory Intelligence Agents for Deep Research
Share to email
Share to Facebook
Share to X
Share to email
Share to Facebook
Share to X
Or
Memory Intelligence Agents for Deep Research
This episode explores a 2026 paper on Memory Intelligence Agent (MIA), a deep research agent designed to move beyond simply storing and retrieving raw past trajectories. It breaks down the paper’s core idea of combining non-parametric memory—an external bank of compressed search experiences—with parametric memory in the planner, so the system can reuse past investigations more efficiently as tasks grow longer and more complex. The discussion highlights why current agent memory systems often scale poorly, becoming expensive, noisy, and cluttered, and examines MIA’s proposed Manager-Planner-Executor architecture as a way to separate memory management, planning, and tool-based execution. Listeners interested in AI agents will find it compelling for its concrete attempt to improve long-horizon research performance through memory compression, test-time self-improvement, and more structured learning.
Sources:
1. Memory Intelligence Agent — Jingyang Qiao, Weicheng Meng, Yu Cheng, Zhihang Lin, Zhizhong Zhang, Xin Tan, Jingyu Gong, Kun Shao, Yuan Xie, 2026
http://arxiv.org/abs/2604.04503
2. Generative Agents: Interactive Simulacra of Human Behavior — Joon Sung Park, Joseph O'Brien, Carrie Cai, Meredith Ringel Morris, Percy Liang, Michael S. Bernstein, 2023
https://scholar.google.com/scholar?q=Generative+Agents:+Interactive+Simulacra+of+Human+Behavior
3. Voyager: An Open-Ended Embodied Agent with Large Language Models — Guanzhi Wang, Yuqi Xie, Hang Yin, Zhenjie Pei, et al., 2023
https://scholar.google.com/scholar?q=Voyager:+An+Open-Ended+Embodied+Agent+with+Large+Language+Models
4. Reflexion: Language Agents with Verbal Reinforcement Learning — Noah Shinn, Federico Cassano, Ashwin Gopinath, et al., 2023
https://scholar.google.com/scholar?q=Reflexion:+Language+Agents+with+Verbal+Reinforcement+Learning
5. MemGPT: Towards LLMs as Operating Systems — Charles Packer, Vivian Fang, Sarah Wooders, Kevin Lin, et al., 2023
https://scholar.google.com/scholar?q=MemGPT:+Towards+LLMs+as+Operating+Systems
6. ReAct: Synergizing Reasoning and Acting in Language Models — Shunyu Yao, Jeffrey Zhao, Dian Yu, et al., 2023
https://scholar.google.com/scholar?q=ReAct:+Synergizing+Reasoning+and+Acting+in+Language+Models
7. Self-Refine: Iterative Refinement with Self-Feedback — Aman Madaan, Niket Tandon, Prakhar Gupta, et al., 2023
https://scholar.google.com/scholar?q=Self-Refine:+Iterative+Refinement+with+Self-Feedback
8. RAG: Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks — Patrick Lewis, Ethan Perez, Aleksandr Piktus, et al., 2020
https://scholar.google.com/scholar?q=RAG:+Retrieval-Augmented+Generation+for+Knowledge-Intensive+NLP+Tasks
9. Toolformer: Language Models Can Teach Themselves to Use Tools — Timo Schick, Jane Dwivedi-Yu, Roberto Dessì, et al., 2023
https://scholar.google.com/scholar?q=Toolformer:+Language+Models+Can+Teach+Themselves+to+Use+Tools
10. LongMem: Scaling Language Models with Long-Term Memory — Yinghan Wang, Yuhang Zang, et al., 2023
https://scholar.google.com/scholar?q=LongMem:+Scaling+Language+Models+with+Long-Term+Memory
11. A-MEM / MemoryBank-style LLM memory papers — Various 2023-2025 authors, 2023-2025
https://scholar.google.com/scholar?q=A-MEM+/+MemoryBank-style+LLM+memory+papers
12. Deciphering the Interplay of Parametric and Non-Parametric Memory in Retrieval-Augmented Language Models — approx. retrieval/RAG interpretability authors, 2024
https://scholar.google.com/scholar?q=Deciphering+the+Interplay+of+Parametric+and+Non-Parametric+Memory+in+Retrieval-Augmented+Language+Models
13. When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories — approx. QA/RAG evaluation authors, 2024
https://scholar.google.com/scholar?q=When+Not+to+Trust+Language+Models:+Investigating+Effectiveness+of+Parametric+and+Non-Parametric+Memories
14. Evo-Memory: Benchmarking LLM Agent Test-Time Learning with Self-Evolving Memory — approx. benchmark authors, 2024
https://scholar.google.com/scholar?q=Evo-Memory:+Benchmarking+LLM+Agent+Test-Time+Learning+with+Self-Evolving+Memory
15. Self-Improving LLM Agents at Test-Time — approx. agent self-improvement authors, 2024
https://scholar.google.com/scholar?q=Self-Improving+LLM+Agents+at+Test-Time
16. Sensi: Learn One Thing at a Time—Curriculum-Based Test-Time Learning for LLM Game Agents — approx. test-time learning / game-agent authors, 2024
https://scholar.google.com/scholar?q=Sensi:+Learn+One+Thing+at+a+Time—Curriculum-Based+Test-Time+Learning+for+LLM+Game+Agents
17. Chain-of-Agents: End-to-End Agent Foundation Models via Multi-Agent Distillation and Agentic RL — approx. multi-agent foundation model authors, 2024
https://scholar.google.com/scholar?q=Chain-of-Agents:+End-to-End+Agent+Foundation+Models+via+Multi-Agent+Distillation+and+Agentic+RL
18. AI Post Transformers: MEMSEARCHER: Reinforcement Learning for LLM Memory Management — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-04-04-memsearcher-reinforcement-learning-for-l-e9ad84.mp3
19. AI Post Transformers: Kosmos AI Scientist for Autonomous Discovery — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-04-04-kosmos-ai-scientist-for-autonomous-disco-311775.mp3
20. AI Post Transformers: MetaClaw: Just Talk and Continual Agent Adaptation — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-03-31-metaclaw-meta-learning-agents-in-the-wil-ab324c.mp3
21. AI Post Transformers: Multi-Agent Tool-Integrated Policy Optimization (MATPO) — Hal Turing & Dr. Ada Shannon, 2025
https://podcast.do-not-panic.com/episodes/multi-agent-tool-integrated-policy-optimization-matpo/
22. AI Post Transformers: MATTRL: Collaborative Test-Time Reinforcement Learning for Multi-Agent Reasoning — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/mattrl-collaborative-test-time-reinforcement-learning-for-multi-agent-reasoning/
23. AI Post Transformers: DeepVerifier: Self-Evolving Research Agents via Rubric-Guided Verification — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/deepverifier-self-evolving-research-agents-via-rubric-guided-verification/
24. AI Post Transformers: Mem0: Scalable Long-Term Memory for AI Agents — Hal Turing & Dr. Ada Shannon, 2025
https://podcast.do-not-panic.com/episodes/mem0-scalable-long-term-memory-for-ai-agents/
25. AI Post Transformers: Doc-to-LoRA: Internalizing Context as LoRA — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-03-29-doc-to-lora-internalizing-context-as-lor-8dd5ec.mp3
26. AI Post Transformers: DeepSeek Engram: Scaling Large Language Models via Conditional Memory Lookup — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/deepseek-engram-scaling-large-language-models-via-conditional-memory-lookup/
Interactive Visualization: Memory Intelligence Agents for Deep Research
View all episodes
By mcgrofThis episode explores a 2026 paper on Memory Intelligence Agent (MIA), a deep research agent designed to move beyond simply storing and retrieving raw past trajectories. It breaks down the paper’s core idea of combining non-parametric memory—an external bank of compressed search experiences—with parametric memory in the planner, so the system can reuse past investigations more efficiently as tasks grow longer and more complex. The discussion highlights why current agent memory systems often scale poorly, becoming expensive, noisy, and cluttered, and examines MIA’s proposed Manager-Planner-Executor architecture as a way to separate memory management, planning, and tool-based execution. Listeners interested in AI agents will find it compelling for its concrete attempt to improve long-horizon research performance through memory compression, test-time self-improvement, and more structured learning.
Sources:
1. Memory Intelligence Agent — Jingyang Qiao, Weicheng Meng, Yu Cheng, Zhihang Lin, Zhizhong Zhang, Xin Tan, Jingyu Gong, Kun Shao, Yuan Xie, 2026
http://arxiv.org/abs/2604.04503
2. Generative Agents: Interactive Simulacra of Human Behavior — Joon Sung Park, Joseph O'Brien, Carrie Cai, Meredith Ringel Morris, Percy Liang, Michael S. Bernstein, 2023
https://scholar.google.com/scholar?q=Generative+Agents:+Interactive+Simulacra+of+Human+Behavior
3. Voyager: An Open-Ended Embodied Agent with Large Language Models — Guanzhi Wang, Yuqi Xie, Hang Yin, Zhenjie Pei, et al., 2023
https://scholar.google.com/scholar?q=Voyager:+An+Open-Ended+Embodied+Agent+with+Large+Language+Models
4. Reflexion: Language Agents with Verbal Reinforcement Learning — Noah Shinn, Federico Cassano, Ashwin Gopinath, et al., 2023
https://scholar.google.com/scholar?q=Reflexion:+Language+Agents+with+Verbal+Reinforcement+Learning
5. MemGPT: Towards LLMs as Operating Systems — Charles Packer, Vivian Fang, Sarah Wooders, Kevin Lin, et al., 2023
https://scholar.google.com/scholar?q=MemGPT:+Towards+LLMs+as+Operating+Systems
6. ReAct: Synergizing Reasoning and Acting in Language Models — Shunyu Yao, Jeffrey Zhao, Dian Yu, et al., 2023
https://scholar.google.com/scholar?q=ReAct:+Synergizing+Reasoning+and+Acting+in+Language+Models
7. Self-Refine: Iterative Refinement with Self-Feedback — Aman Madaan, Niket Tandon, Prakhar Gupta, et al., 2023
https://scholar.google.com/scholar?q=Self-Refine:+Iterative+Refinement+with+Self-Feedback
8. RAG: Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks — Patrick Lewis, Ethan Perez, Aleksandr Piktus, et al., 2020
https://scholar.google.com/scholar?q=RAG:+Retrieval-Augmented+Generation+for+Knowledge-Intensive+NLP+Tasks
9. Toolformer: Language Models Can Teach Themselves to Use Tools — Timo Schick, Jane Dwivedi-Yu, Roberto Dessì, et al., 2023
https://scholar.google.com/scholar?q=Toolformer:+Language+Models+Can+Teach+Themselves+to+Use+Tools
10. LongMem: Scaling Language Models with Long-Term Memory — Yinghan Wang, Yuhang Zang, et al., 2023
https://scholar.google.com/scholar?q=LongMem:+Scaling+Language+Models+with+Long-Term+Memory
11. A-MEM / MemoryBank-style LLM memory papers — Various 2023-2025 authors, 2023-2025
https://scholar.google.com/scholar?q=A-MEM+/+MemoryBank-style+LLM+memory+papers
12. Deciphering the Interplay of Parametric and Non-Parametric Memory in Retrieval-Augmented Language Models — approx. retrieval/RAG interpretability authors, 2024
https://scholar.google.com/scholar?q=Deciphering+the+Interplay+of+Parametric+and+Non-Parametric+Memory+in+Retrieval-Augmented+Language+Models
13. When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memories — approx. QA/RAG evaluation authors, 2024
https://scholar.google.com/scholar?q=When+Not+to+Trust+Language+Models:+Investigating+Effectiveness+of+Parametric+and+Non-Parametric+Memories
14. Evo-Memory: Benchmarking LLM Agent Test-Time Learning with Self-Evolving Memory — approx. benchmark authors, 2024
https://scholar.google.com/scholar?q=Evo-Memory:+Benchmarking+LLM+Agent+Test-Time+Learning+with+Self-Evolving+Memory
15. Self-Improving LLM Agents at Test-Time — approx. agent self-improvement authors, 2024
https://scholar.google.com/scholar?q=Self-Improving+LLM+Agents+at+Test-Time
16. Sensi: Learn One Thing at a Time—Curriculum-Based Test-Time Learning for LLM Game Agents — approx. test-time learning / game-agent authors, 2024
https://scholar.google.com/scholar?q=Sensi:+Learn+One+Thing+at+a+Time—Curriculum-Based+Test-Time+Learning+for+LLM+Game+Agents
17. Chain-of-Agents: End-to-End Agent Foundation Models via Multi-Agent Distillation and Agentic RL — approx. multi-agent foundation model authors, 2024
https://scholar.google.com/scholar?q=Chain-of-Agents:+End-to-End+Agent+Foundation+Models+via+Multi-Agent+Distillation+and+Agentic+RL
18. AI Post Transformers: MEMSEARCHER: Reinforcement Learning for LLM Memory Management — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-04-04-memsearcher-reinforcement-learning-for-l-e9ad84.mp3
19. AI Post Transformers: Kosmos AI Scientist for Autonomous Discovery — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-04-04-kosmos-ai-scientist-for-autonomous-disco-311775.mp3
20. AI Post Transformers: MetaClaw: Just Talk and Continual Agent Adaptation — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-03-31-metaclaw-meta-learning-agents-in-the-wil-ab324c.mp3
21. AI Post Transformers: Multi-Agent Tool-Integrated Policy Optimization (MATPO) — Hal Turing & Dr. Ada Shannon, 2025
https://podcast.do-not-panic.com/episodes/multi-agent-tool-integrated-policy-optimization-matpo/
22. AI Post Transformers: MATTRL: Collaborative Test-Time Reinforcement Learning for Multi-Agent Reasoning — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/mattrl-collaborative-test-time-reinforcement-learning-for-multi-agent-reasoning/
23. AI Post Transformers: DeepVerifier: Self-Evolving Research Agents via Rubric-Guided Verification — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/deepverifier-self-evolving-research-agents-via-rubric-guided-verification/
24. AI Post Transformers: Mem0: Scalable Long-Term Memory for AI Agents — Hal Turing & Dr. Ada Shannon, 2025
https://podcast.do-not-panic.com/episodes/mem0-scalable-long-term-memory-for-ai-agents/
25. AI Post Transformers: Doc-to-LoRA: Internalizing Context as LoRA — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/2026-03-29-doc-to-lora-internalizing-context-as-lor-8dd5ec.mp3
26. AI Post Transformers: DeepSeek Engram: Scaling Large Language Models via Conditional Memory Lookup — Hal Turing & Dr. Ada Shannon, 2026
https://podcast.do-not-panic.com/episodes/deepseek-engram-scaling-large-language-models-via-conditional-memory-lookup/
Interactive Visualization: Memory Intelligence Agents for Deep Research