{"paper":{"title":"Retrieval-Augmented LLM Agents: Learning to Learn from Experience","arxiv_id":"2603.18272","primary_cat":"cs.AI","submitted_at":"2026-03-18T20:45:04+00:00","version_compiled":"v2"},"notice":"Every quoted sentence is the authors' text from the arXiv listing abstract. Labels and probabilities are the outputs of classifier-1.13.0 on the question bank abstract-p1-20260923, exactly as returned. No sentence in this record was written by a language model.","schema":"pith.compiled-record/v0","status":"compiled","abstract":{"source":"arXiv listing abstract v2","sentences":[{"i":0,"p":0.92,"id":"s0","own":false,"role":"background","text":"While large language models (LLMs) have advanced the development of general-purpose agents, robust generalization to unseen tasks remains challenging.","hedge":{"p":0.99,"label":"asserted"},"own_p":0.04,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.07,"role_probabilities":{"gap":0.08,"method":0.0,"result":0.0,"background":0.92,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":1,"p":0.78,"id":"s1","own":false,"role":"background","text":"Two common approaches are supervised fine-tuning and training-free memory-augmented generation using retrieved experience; yet both have limitations: fine-tuning often fails to extrapolate to new tasks, while experience retrieval often underperforms compared to supervised baselines.","hedge":{"p":0.92,"label":"asserted"},"own_p":0.06,"main_claim_p":0.03,"quantitative":true,"quantitative_p":0.78,"role_probabilities":{"gap":0.21,"method":0.0,"result":0.0,"background":0.78,"limitation":0.01,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":2,"p":0.99,"id":"s2","own":true,"role":"contribution","text":"In this work, we combine these approaches and study how retrieval-augmented LLM agents can learn to use retrieved trajectories in-context.","hedge":{"p":0.65,"label":"asserted"},"own_p":0.63,"main_claim_p":0.05,"quantitative":false,"quantitative_p":0.03,"role_probabilities":{"gap":0.0,"method":0.01,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.99,"interpretation":0.0}},{"i":3,"p":0.54,"id":"s3","own":true,"role":"result","text":"First, we establish a strong LoRA fine-tuning baseline that outperforms several state-of-the-art agent training pipelines.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.95,"main_claim_p":0.04,"quantitative":true,"quantitative_p":0.78,"role_probabilities":{"gap":0.0,"method":0.23,"result":0.54,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.23,"interpretation":0.0}},{"i":4,"p":0.52,"id":"s4","own":true,"role":"method","text":"Second, we analyze key design choices for experience retrieval, including storage, querying, and trajectory selection.","hedge":{"p":0.65,"label":"asserted"},"own_p":0.89,"main_claim_p":0.02,"quantitative":false,"quantitative_p":0.08,"role_probabilities":{"gap":0.0,"method":0.52,"result":0.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.48,"interpretation":0.0}},{"i":5,"p":0.96,"id":"s5","own":true,"role":"result","text":"We then integrate experience retrieval directly into the fine-tuning process, finding that this substantially improves generalization to unseen tasks.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.97,"main_claim_p":0.27,"quantitative":false,"quantitative_p":0.16,"role_probabilities":{"gap":0.0,"method":0.02,"result":0.96,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.02,"interpretation":0.0}},{"i":6,"p":1.0,"id":"s6","own":true,"role":"result","text":"Finally, we show that these gains often persist with imperfect experience and, even when agents reuse their own failed attempts without test-time parameter updates.","hedge":{"p":0.98,"label":"asserted"},"own_p":0.95,"main_claim_p":0.12,"quantitative":false,"quantitative_p":0.09,"role_probabilities":{"gap":0.0,"method":0.0,"result":1.0,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.0,"interpretation":0.0}},{"i":7,"p":0.94,"id":"s7","own":true,"role":"interpretation","text":"Overall, our results establish simple episodic retrieval as a strong foundation for agent memory and retrieval-aware fine-tuning as a practical and effective framework for building agents that learn to learn from experience.","hedge":{"p":1.0,"label":"asserted"},"own_p":0.91,"main_claim_p":0.96,"quantitative":false,"quantitative_p":0.07,"role_probabilities":{"gap":0.0,"method":0.0,"result":0.03,"background":0.0,"limitation":0.0,"future_work":0.0,"contribution":0.03,"interpretation":0.94}}]},"compiled":{"main_claim":{"p":1.0,"text":"Overall, our results establish simple episodic retrieval as a strong foundation for agent memory and retrieval-aware fine-tuning as a practical and effective framework for building agents that learn to learn from experience.","location":"abstract s7","hedge_strength":{"p":1.0,"label":"asserted"}},"provenance":{"route":{"via":"TypeSafe API","model":"classifier-1.13.0","endpoint":"https://api.typesafe.ai/v1/systemone","provider":"TypeSafe"},"packing":"one request per paper","requests":1,"input_tokens":6273,"cost_usd_reported":0.0},"compiled_at":"2026-09-23","bank_version":"abstract-p1-20260923","model_version":"classifier-1.13.0","abstract_states":{"claims_first":0.05,"states_limitation":0.52,"human_participants":0.04,"controlled_comparison":0.91,"randomized_assignment":0.02,"states_quantitative_result":0.05,"compares_with_prior_results":0.97,"mentions_code_or_data_availability":0.03},"evidential_basis":{"p":0.96,"value":"observation or experiment","probabilities":{"argument":0.0,"literature":0.0,"proof_or_derivation":0.0,"analysis_of_existing_data":0.0,"observation_or_experiment":0.96,"simulation_or_computation":0.04}},"contribution_type":{"p":0.76,"value":"method or tool","probabilities":{"review":0.0,"method_or_tool":0.76,"empirical_study":0.24,"simulation_study":0.0,"theoretical_result":0.0,"dataset_or_benchmark":0.0,"position_or_commentary":0.0}},"contribution_statement":{"p":0.59,"text":"In this work, we combine these approaches and study how retrieval-augmented LLM agents can learn to use retrieved trajectories in-context.","location":"abstract s2"}},"coverage":{"tier":"abstract","levels_run":["P abstract"],"not_examined":["full text","figures and tables","references and citing sentences","relations to other papers"],"sources_read":["arXiv listing abstract v2"],"levels_not_run":["S sections","T body sentences","C citations","R cross-paper relations"],"sentences_examined":8,"sentences_in_abstract":8},"measured_accuracy":{"url":"https://pith.science/accuracy","measured_on":"2026-10-02","method":"two independent model families labelled a fresh random sample blind; the figure is how often the published label matches their consensus","sentence_role_agreement":0.92,"main_claim_agreement":0.908,"typed_citation_agreement":0.933,"typed_citation_agreement_95_interval":[0.917,0.945],"review_comments_accurate":0.758,"review_comments_accurate_95_interval":[0.684,0.82],"review_caught_admitted_error_found_or_related":0.448,"review_comments_judged":{"judged_by":["gpt-5.6-luna","grok-4.7"],"judged_against":"Pith's own text copy of each paper: text extracted from the PDF, the first 90,000 characters","measured_on":"2026-10-04","accurate":113,"decided":149,"sampled_reviews":200,"split_by_lane":false,"by_lane":"https://pith.science/accuracy#by-lane"}}}