Memory Transfer Learning: How Memories are Transferred Across Domains in Coding Agents

Kangsan Kim¹, Minki Kang¹, Taeil Kim¹, Yanlai Yang², Mengye Ren^2†, Sung Ju Hwang^1,3†

¹KAIST    ²New York University    ³DeepAuto.ai    ^†Equal advising

Correspondence: kangsan.kim@kaist.ac.kr

---

TL;DR

We investigate cross-domain memory transfer for coding agents and show that leveraging a unified memory pool from heterogeneous benchmarks improves average performance by 3.7%. Abstraction is the key: high-level insights generalize across domains while low-level traces induce negative transfer.

Overview

Existing self-evolving coding agents restrict memory usage to the same benchmark. We propose Memory Transfer Learning (MTL), which leverages a unified memory pool from heterogeneous domains, and show it consistently outperforms domain-restricted approaches.

(A) Memory-less agents cannot reflect on past experience. (B) Self-evolving agents leverage memory but only within a single domain. (C) MTL leverages a unified memory pool from heterogeneous coding tasks. (D) MTL (hatched bars) consistently outperforms self-evolving agents across all four memory formats.

Research Questions

• RQ1: Does memory from heterogeneous domains improve the performance of coding agents?
• RQ2: Why do transferred memories yield benefits across different domains?
• RQ3: Which factors in memory transfer learning most influence transfer effectiveness?

Method

Memory Representations

We construct four types of memory representations spanning a spectrum from concrete low-level traces to abstract high-level insights:

Type	Description
Trajectory	Concatenates all agent commands and execution results. Contains full task-solving detail including failed steps.
Workflow	Extracts a reusable goal-oriented workflow — a goal statement plus a subset of meaningful actions.
Summary	Prompts an LLM to summarize the task, environment, actions, and an analysis of success or failure.
Insight	Generalizable principles written to be task-agnostic for effective cross-domain transfer.

Memory Retrieval Pipeline

1. Memory Generation — Run the agent across all benchmarks. Use an LLM judge to assess success/failure, then generate all four memory types from each trajectory.
2. Pool Construction — Merge memories from all benchmarks except the target. Index each memory using text-embedding-3-small for fast retrieval.
3. Retrieval & Inference — For each query, retrieve the top-3 most similar memories and prepend them to the coding agent's system prompt before inference.

Results

Main Results (Pass@3 across 6 Benchmarks)

Method	LiveCodeBenchv6	Aider-Polyglot	SWEBench-Verified	TerminalBench2	ReplicationBench	MLGym-Bench	Avg.
GPT-5-mini
Zero-Shot	0.910	0.470	0.730	0.315	0.111	0.667	0.523
MTL (T)	0.940	0.490	0.770	0.270	0.122	0.583	0.534
MTL (W)	0.920	0.470	0.770	0.348	0.111	0.583	0.538
MTL (S)	0.930	0.460	0.760	0.371	0.133	0.667	0.546
MTL (I)	0.930	0.470	0.770	0.360	0.189	0.750	0.560
Δ	+2.0%	0.0%	+4.0%	+4.5%	+7.8%	+8.3%	+3.7%

Comparison with Self-Evolving Baselines

MTL outperforms ReasoningBank (+2.9%) and AgentKB (+1.7%) with only 431 memories — far fewer than AgentKB's 5,899 memories.

Method	#Memories	LiveCodeBenchv6	SWEBench-Verified	ReplicationBench	Avg.
Zero-Shot	—	0.910	0.730	0.111	0.584
ReasoningBank	97	0.920	0.750	0.133	0.601
AgentKB	5,899	0.920	0.720	0.200	0.613
MTL (Ours)	431	0.930	0.770	0.189	0.630

Key Findings

• Finding 1: MTL significantly improves coding agent performance and outperforms self-evolving methods in both effectiveness and efficiency.
• Finding 2: The primary form of transferable knowledge is meta-memory encoding procedural and behavioral guidance — not domain-specific code.
• Finding 3: More abstract and generalized memory representations yield higher transfer effectiveness by avoiding brittle implementation anchoring.
• Finding 4: Negative memory transfer arises from domain-mismatched misleading anchors, false validation signals, and misapplied procedural reuse.
• Finding 5: MTL effectiveness scales with the size of the memory pool and the number of source domains.
• Finding 6: Memory can be transferred across different models; self-generated memories yield the best performance, but cross-model transfer consistently beats zero-shot.
• Finding 7: Cross-domain memory retrieval is inherently challenging; static retrieval methods fail to generalize in heterogeneous agentic settings.

Code

Coming Soon

Acknowledgements

Our work builds upon Harbor and Mini-SWE-Agent. We thank the authors for releasing their code.

BibTeX

@misc{kim2026memorytransferlearningmemories,
  title={Memory Transfer Learning: How Memories are Transferred Across Domains in Coding Agents}, 
  author={Kangsan Kim and Minki Kang and Taeil Kim and Yanlai Yang and Mengye Ren and Sung Ju Hwang},
  year={2026},
  eprint={2604.14004},
  archivePrefix={arXiv},
  primaryClass={cs.AI},
  url={https://arxiv.org/abs/2604.14004}, 
}