# VLM Agents Generate Their Own Memories: Distilling Experience into Embodied Programs of Thought

arXiv: https://arxiv.org/abs/2406.14596 · HTML: https://mlsanity.com/paper/2406.14596/

## Key facts

- Cited by 55 papers in the map (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 35 of its 96 references that are in the map.
- First posted on arXiv 20 Jun 2024.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [Embodied and 3D vision-language models](https://mlsanity.com/mllm/embodied-and-3d-vision-language-models/)
- Retrieval-augmented generation / [Agentic RAG, agents and memory](https://mlsanity.com/rag/agentic-rag-agents-and-memory/)
- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 30 of 35)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [LoRA: Low-Rank Adaptation of Large Language Models](https://mlsanity.com/paper/2106.09685/) (2021)
- [Chain-of-Thought Prompting Elicits Reasoning in Large Language Models](https://mlsanity.com/paper/2201.11903/) (2022)
- [Self-Consistency Improves Chain of Thought Reasoning in Language Models](https://mlsanity.com/paper/2203.11171/) (2022)
- [Reflexion: Language Agents with Verbal Reinforcement Learning](https://mlsanity.com/paper/2303.11366/) (2023)
- [Code as Policies: Language Model Programs for Embodied Control](https://mlsanity.com/paper/2209.07753/) (2022)
- [Reproducible scaling laws for contrastive language-image learning](https://mlsanity.com/paper/2212.07143/) (2022)
- [SpatialVLM: Endowing Vision-Language Models with Spatial Reasoning Capabilities](https://mlsanity.com/paper/2401.12168/) (2024)
- [ZoeDepth: Zero-shot Transfer by Combining Relative and Metric Depth](https://mlsanity.com/paper/2302.12288/) (2023)
- [Eureka: Human-Level Reward Design via Coding Large Language Models](https://mlsanity.com/paper/2310.12931/) (2023)
- [OpenFlamingo: An Open-Source Framework for Training Large Autoregressive Vision-Language Models](https://mlsanity.com/paper/2308.01390/) (2023)
- [GPT-4V(ision) is a Generalist Web Agent, if Grounded](https://mlsanity.com/paper/2401.01614/) (2024)
- [V*: Guided Visual Search as a Core Mechanism in Multimodal LLMs](https://mlsanity.com/paper/2312.14135/) (2023)
- [Describe, Explain, Plan and Select: Interactive Planning with Large Language Models Enables Open-World Multi-Task Agents](https://mlsanity.com/paper/2302.01560/) (2023)
- [Set-of-Mark Prompting Unleashes Extraordinary Visual Grounding in GPT-4V](https://mlsanity.com/paper/2310.11441/) (2023)
- [Reinforcement Learning on Web Interfaces Using Workflow-Guided Exploration](https://mlsanity.com/paper/1802.08802/) (2018)
- [Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual Concepts](https://mlsanity.com/paper/2111.08276/) (2021)
- [REFLECT: Summarizing Robot Experiences for Failure Explanation and Correction](https://mlsanity.com/paper/2306.15724/) (2023)
- [JARVIS-1: Open-World Multi-task Agents with Memory-Augmented Multimodal Language Models](https://mlsanity.com/paper/2311.05997/) (2023)
- [Using Natural Language for Reward Shaping in Reinforcement Learning](https://mlsanity.com/paper/1903.02020/) (2019)
- [HomeRobot: Open-Vocabulary Mobile Manipulation](https://mlsanity.com/paper/2306.11565/) (2023)
- [Motif: Intrinsic Motivation from Artificial Intelligence Feedback](https://mlsanity.com/paper/2310.00166/) (2023)
- [Learning to Understand Goal Specifications by Modelling Reward](https://mlsanity.com/paper/1806.01946/) (2018)
- [From Language to Goals: Inverse Reinforcement Learning for Vision-Based Instruction Following](https://mlsanity.com/paper/1902.07742/) (2019)
- [Navigating to Objects Specified by Images](https://mlsanity.com/paper/2304.01192/) (2023)
- [VisualWebArena: Evaluating Multimodal Agents on Realistic Visual Web Tasks](https://mlsanity.com/paper/2401.13649/) (2024)
- [Model-Based Runtime Monitoring with Interactive Imitation Learning](https://mlsanity.com/paper/2310.17552/) (2023)
- [LLF-Bench: Benchmark for Interactive Learning from Language Feedback](https://mlsanity.com/paper/2312.06853/) (2023)
- [Think, Act, and Ask: Open-World Interactive Personalized Robot Navigation](https://mlsanity.com/paper/2310.07968/) (2023)
- [Interactive Robot Learning from Verbal Correction](https://mlsanity.com/paper/2310.17555/) (2023)

## Built on by (most cited first, 30 of 55)

- [The BrowserGym Ecosystem for Web Agent Research](https://mlsanity.com/paper/2412.05467/) (2024)
- [Inducing Programmatic Skills for Agentic Tasks](https://mlsanity.com/paper/2504.06821/) (2025)
- [Lifelong Learning of Large Language Model based Agents: A Roadmap](https://mlsanity.com/paper/2501.07278/) (2025)
- [Grounded Reinforcement Learning for Visual Reasoning](https://mlsanity.com/paper/2505.23678/) (2025)
- [Agent-R: Training Language Model Agents to Reflect via Iterative Self-Training](https://mlsanity.com/paper/2501.11425/) (2025)
- [A Survey of LLM-Driven AI Agent Communication: Protocols, Security Risks, and Defense Countermeasures](https://mlsanity.com/paper/2506.19676/) (2025)
- [AGENTSAFE: Benchmarking the Safety of Embodied Agents on Hazardous Instructions](https://mlsanity.com/paper/2506.14697/) (2025)
- [Visual Large Language Models for Generalized and Specialized Applications](https://mlsanity.com/paper/2501.02765/) (2025)
- [GUI-explorer: Autonomous Exploration and Mining of Transition-aware Knowledge for GUI Agent](https://mlsanity.com/paper/2505.16827/) (2025)
- [A Survey of Multimodal Hallucination Evaluation and Detection](https://mlsanity.com/paper/2507.19024/) (2025)
- [WebXSkill: Skill Learning for Autonomous Web Agents](https://mlsanity.com/paper/2604.13318/) (2026)
- [MME-VideoOCR: Evaluating OCR-Based Capabilities of Multimodal LLMs in Video Scenarios](https://mlsanity.com/paper/2505.21333/) (2025)
- [WALT: Web Agents that Learn Tools](https://mlsanity.com/paper/2510.01524/) (2025)
- [SkillDAG: Self-Evolving Typed Skill Graphs for LLM Skill Selection at Scale](https://mlsanity.com/paper/2606.03056/) (2026)
- [Let's Think in Two Steps: Mitigating Agreement Bias in MLLMs with Self-Grounded Verification](https://mlsanity.com/paper/2507.11662/) (2025)
- [IAG: Input-aware Backdoor Attack on VLM-based Visual Grounding](https://mlsanity.com/paper/2508.09456/) (2025)
- [MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents](https://mlsanity.com/paper/2605.18652/) (2026)
- [LENS: Multi-level Evaluation of Multimodal Reasoning with Large Language Models](https://mlsanity.com/paper/2505.15616/) (2025)
- [Unleashing Embodied Task Planning Ability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2506.23127/) (2025)
- [Memento No More: Coaching AI Agents to Master Multiple Tasks via Hints Internalization](https://mlsanity.com/paper/2502.01562/) (2025)
- [Position: Scaling LLM Agents Requires Asymptotic Analysis with LLM Primitives](https://mlsanity.com/paper/2502.04358/) (2025)
- [RoboAgent: Chaining Basic Capabilities for Embodied Task Planning](https://mlsanity.com/paper/2604.07774/) (2026)
- [Building LLM Agents by Incorporating Insights from Computer Systems](https://mlsanity.com/paper/2504.04485/) (2025)
- [Recon-Act: A Self-Evolving Multi-Agent Browser-Use System via Web Reconnaissance, Tool Generation, and Task Execution](https://mlsanity.com/paper/2509.21072/) (2025)
- [Same or Not? Enhancing Visual Perception in Vision-Language Models](https://mlsanity.com/paper/2512.23592/) (2025)
- [A Review of 3D Object Detection with Vision-Language Models](https://mlsanity.com/paper/2504.18738/) (2025)
- [ReEXplore: Improving MLLMs for Embodied Exploration with Contextualized Retrospective Experience Replay](https://mlsanity.com/paper/2511.19033/) (2025)
- [No Labels, No Problem: Training Visual Reasoners with Multimodal Verifiers](https://mlsanity.com/paper/2512.08889/) (2025)
- [M2: Dual-Memory Augmentation for Long-Horizon Web Agents via Trajectory Summarization and Insight Retrieval](https://mlsanity.com/paper/2603.00503/) (2026)
- [MemToolAgent: Leveraging Memory for Tool Using Agents Based on Environment and User Feedback](https://mlsanity.com/paper/2606.07909/) (2026)

## Abstract (arXiv)

Large-scale generative language and vision-language models (LLMs and VLMs) excel in few-shot learning but require high-quality demonstrations. We propose In-Context Abstraction Learning (ICAL), enabling VLM agents to transform suboptimal trajectories into high-quality training data through self-reflection and human feedback. Given imperfect task demonstrations, a VLM abstracts trajectories into generalized strategies and action annotations by correcting inefficiencies and annotating cognitive abstractions: causal relationships, object state changes, temporal subgoals, and task-relevant visual elements. These annotations are iteratively refined through human feedback during execution in similar environments. The resulting examples significantly improve decision-making when used for retrieval-augmented generation or fine-tuning. As the agent's example library grows, it becomes more efficient at abstracting new examples, requiring less human feedback and fewer environment interactions. ICAL achieves state-of-the-art results across multiple benchmarks. In TEACh dialogue-based instruction following, combining fine-tuning and retrieval on ICAL examples outperforms raw human demonstrations and expert examples by 17.5% in goal-condition success. In VisualWebArena, retrieval-augmented GPT-4V with ICAL improves task success 1.6x, while fine-tuned Qwen2-VL achieves 2.8x improvement over the base model. In Ego4D action forecasting, we surpass few-shot GPT-4V and remain competitive with supervised models. Our approach scales 2x better than raw demonstrations and significantly reduces manual prompt engineering requirements.