XferAgent: Cross-Agent Experience Transfer for Self-Improving LLM Agents

Weihao Li; Jun Bai; Shijie Chen; Qincheng Lu; Xinyu Wang; Ziyang Song

cs.MAOct 14, 2026

XferAgent: Cross-Agent Experience Transfer for Self-Improving LLM Agents

Authors: Weihao Li, Jun Bai, Shijie Chen, Qincheng Lu, Xinyu Wang, Ziyang Song

Organizations: McCormick School of Engineering, Northwestern University · School of Computer Science, McGill University · Mila – Quebec AI Institute · School of Electrical Engineering and Computer Science, Ohio University

Abstract

Large language model (LLM) agents can reuse experience from prior task executions to improve performance on subsequent tasks. Although existing methods enable cross-agent experience sharing, differences in agents' execution histories and task requirements make shared experiences difficult to reuse effectively. We propose XferAgent, a framework for cross-agent experience transfer that organizes reusable experience in local and shared experience graphs. XferAgent reconciles compatible contributions from heterogeneous agents while avoiding unreliable merges. For each new task, it selects shared experiences and integrates them with the target agent's local experience to support task execution. Experiments across diverse task domains show improvements over representative memory-based methods and effective experience transfer between heterogeneous agent frameworks. Cumulative task success also increases over successive task passes as agents update and reuse their experiences.

Figures & tables

Figure 1: Overview of XferAgent. Agents construct local experience graphs by extracting experiences from successful and failed task executions. Cross-Agent Experience Reconciliation incorporates contributed experiences into a shared experience graph. Task-Conditioned Experience Routing selects task-relevant anchors and constructs a routed experience subgraph. Local–Shared Experience Integration combines local and shared candidates into a temporary fused graph and ranks experiences for task execution, followed by evaluation-based experience updates.

	GAIA		ALFWorld	MBPP
Method	Qwen3-Next-80B A3B-Instruct	GPT-5.6 Luna	Qwen3.8-27B	Llama-3.1-8B Instruct
No Memory	27.57	48.17	84.7	50.0
AgentKB	28.90	46.51	88.3	50.0
ReasoningBank	26.91	46.18	85.0	46.7
ExpGraph	27.24	47.51	88.3	48.3
Mem0	25.91	44.85	84.7	45.0

Table 1: Overall performance across GAIA, ALFWorld, and MBPP (%). Metrics are accuracy, task success rate, and program pass rate, respectively. ALFWorld and MBPP follow benchmark-specific single-run protocols (Appendix C ). Bold indicates the best result in each column.

Figure 2: Cumulative task performance through experience evolution on GAIA validation (165 tasks). Passes 1–3 report cumulative task success across repeated executions. The independent No Memory baseline is connected to Pass 1 by dashed segments for visual comparison. Overall accuracy is computed over all 165 tasks.

Figure 3: Cross-framework experience transfer with 2, 4, and 8 agents on 60 held-out GAIA validation questions. Gray bars represent the 2-agent baseline, while light and dark red segments indicate accuracy gains from 2 to 4 and 4 to 8 agents, respectively. Average denotes the unweighted mean of accuracy across the three difficulty levels.

Variant	Overall (%)	$\Delta$ (pp)
XferAgent (Full)	52.16	–
Local Experience Only	46.84	-5.32
Shared Experience Only	50.83	-1.33
w/o Outcome-Based Ranking	48.17	-3.99
w/o Connectivity Constraint	46.84	-5.32

Table 2: Ablation study of XferAgent on GAIA (301 test tasks) using GPT-5.6 Luna. Overall accuracy is reported in percentages, and

\Delta

denotes the change relative to the full system in percentage points.

Parameter	Value	Overall (%)	$\Delta$ (pp)
$B$	8	51.11	-6.67
	24 (default)	57.78	–
	40	46.67	-11.11
$K$	1	51.11	-6.67
	3 (default)	57.78	–
	5	62.22	+4.44

Table 3: Parameter sensitivity of XferAgent on 45 held-out GAIA validation tasks using GPT-5.6 Luna. One parameter is varied at a time, with others fixed at their default values.

\Delta

denotes the accuracy change relative to the default in percentage points.

Appendix figures & tables5 assets

Supplementary material from the paper’s appendix.

Appendix

Qwen3-Next-80B-A3B-Instruct
Method	Overall	L1	L2	L3
No Memory	27.57	41.94	26.42	4.08
AgentKB	28.90	49.46	22.64	10.20
ReasoningBank	26.91	43.01	23.27	8.16
ExpGraph	27.24	44.09	24.53	4.08
Mem0	25.91	41.94	23.27	4.08

Appendix

Table 4: GAIA Test accuracy (%) by difficulty. Bold marks the best score in each column within each model block, including ties.

Setting	ALFWorld	MBPP	Cross-framework GAIA
Backbone	Qwen3.8-27B (Tinker)	Llama-3.1-8B-Instruct	GPT-5.6 Luna
Construction data	90 training trajectories	265 cached training trajectories	20 questions per agent from a 105-question source pool
Evaluation cohort	274 tasks: 140 Seen, 134 Unseen	60 development tasks	60 tasks: 18 / 29 / 13 at Levels 1–3
Agent arrangement	Three logical agents	Three logical agents	Nested 2 / 4 / 8 contributors; one fixed receiver
Shared budget $B$ / final limit $K$	8 / 10	8 / 10	24 / 3
Memory allowance	2,400 tokens	2,400 tokens (character estimate)	12,000 tokens

Appendix

Table 5: Settings for the additional evaluations. Memory content and outcome statistics are frozen during evaluation. Memory allowances specify the token limits for retrieved context.

Method	Benchmark-specific adaptation
AgentKB	Flat workflow entries; bounded local and cross-client similarity retrieval.
ReasoningBank	Dense retrieval over cached training experiences.
ExpGraph	Graph retrieval over cached training experiences, frozen during evaluation.
Mem0	Source-preserving procedural-memory updates in a persistent store.
LightMem	Short-term grouping and offline consolidation with source text retained.

Appendix

Table 6: Memory adapters used for MBPP.

Setting	Value
Shared node budget / final limit	8 / 10
Memory allowance (tokens)	2,400
Experience-text limit (characters)	5,000
Reconciliation threshold / margin	0.92 / 0
Structural-overlap threshold	0.50
Anchor limit / task-match threshold	10 / 0.08

Appendix

Table 7: Key settings of the ALFWorld XferAgent adapter.

Agents	SmolAgents	OWL	Total instances
2	1	1	40
4	2	2	80
8	4	4	160

Appendix

Table 8: Nested contributor groups for cross-framework transfer. Each contributor builds memory from 20 source questions. Totals count agent–question instances; questions may overlap across agents.

Explore similar work

LifeMem: Enabling Lifelong Experience Reuse for LLM Agents
Sep 11, 2026Yuli Qiu, Yutong Li, Wei Su +5Continual Learning for LLM Agents LLM Agent Skill Learning
School of Computer Science and Technology, Beijing Institute of Technology · School of Computer Science and Engineering, Beihang University · Research Center for Social Computing and Interactive Robotics, Harbin Institute of Technology +1
ExpGraph: Model-Agnostic Experience Learning with Graph-Structured Memory for LLM Agents
May 29, 2026Tao Feng, Chongrui Ye, Tianyang Luo +8Continual Learning for LLM Agents LLM Agent Skill Learning
University of Illinois Urbana-Champaign · 2Nanyang Technological University · 3Meta Monetization AI
Experiential Reflective Learning for Self-Improving LLM Agents
Mar 25, 2026Marc-Antoine Allard, Arnaud Teinturier, Victor Xing +1Self-Improving Agents LLM Agent Training
Illuin Technology

Sep 11, 2026J/K move · Enter open · S save

LifeMem: Enabling Lifelong Experience Reuse for LLM Agents

Yuli Qiu, Yutong Li, Wei Su, Zeming Liu+4

School of Computer Science and Technology, Beijing Institute of Technology · School of Computer Science and Engineering, Beihang University · Research Center for Social Computing and Interactive Robotics, Harbin Institute of Technology+1

Continual Learning for LLM Agents LLM Agent Skill Learning Transfer Learning Memory-Augmented Language Model Agents

Open paper AskPDF arXiv

Large language model agents are expected to continuously adapt to new tasks and environments over their lifetime by reusing past experience. However, existing memory-based agents struggle to transfer reusable experience across environments and suffer from catastrophic forgetting as experience accumulated. To address these challenges, we propose LifeMem, a lifelong learning framework that enables agents to transfer knowledge across multiple environments. During learning, LifeMem clusters accumulated interaction trajectories based on underlying workflows to extract reusable skills. When solving a new task at inference time, the agent recalls relevant skills and trajectories to guide actions. To validate our method, we conduct experiments across 10 environments and over 13k tasks with 2k newly annotated interaction trajectories. Results show that LifeMem enables effective experience reuse in lifelong learning, achieving both reduced forgetting on learned tasks and superior cross-task transfer. Further analysis reveals that task streaming impacts learning, while consolidating structurally similar trajectories within memory boosts performance.

Abstract

Explore similar work

LifeMem: Enabling Lifelong Experience Reuse for LLM Agents

ExpGraph: Model-Agnostic Experience Learning with Graph-Structured Memory for LLM Agents

Experiential Reflective Learning for Self-Improving LLM Agents