Salvato in:
Dettagli Bibliografici
Autori principali: Shao, Jie-Jing, Yin, Haiyan, Lyu, Yueming, Yu, Xingrui, Guo, Lan-Zhe, Tsang, Ivor, Kwok, James, Li, Yu-Feng
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:https://arxiv.org/abs/2605.01293
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866918477979189248
author Shao, Jie-Jing
Yin, Haiyan
Lyu, Yueming
Yu, Xingrui
Guo, Lan-Zhe
Tsang, Ivor
Kwok, James
Li, Yu-Feng
author_facet Shao, Jie-Jing
Yin, Haiyan
Lyu, Yueming
Yu, Xingrui
Guo, Lan-Zhe
Tsang, Ivor
Kwok, James
Li, Yu-Feng
contents Foundation model-driven agents often struggle with long-horizon planning due to the transient nature of purely prompting-based reasoning. While existing skill induction methods mitigate this by distilling experience into state-blind parameterized scripts, they fail to capture the conditional logic required for robust execution in dynamic environments. In this paper, we propose Neuro-Symbolic Skill Induction (NSI), a framework that lifts interaction traces into modular, \textit{logic-grounded} programs. By synthesizing explicit control flows and dynamic variable binding, NSI empowers agents to discover \textit{when} and \textit{why} to act. This paradigm enables the efficient generalization, allowing agents to induce skills from few-shot examples and flexibly adapt to unseen goals. Experiments on a series of agentic tasks demonstrate that NSI consistently outperforms state-of-the-art baselines, empowering agents to self-evolve into architects of logic-grounded skills.
format Preprint
id arxiv_https___arxiv_org_abs_2605_01293
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Lifting Traces to Logic: Programmatic Skill Induction with Neuro-Symbolic Learning for Long-Horizon Agentic Tasks
Shao, Jie-Jing
Yin, Haiyan
Lyu, Yueming
Yu, Xingrui
Guo, Lan-Zhe
Tsang, Ivor
Kwok, James
Li, Yu-Feng
Artificial Intelligence
Foundation model-driven agents often struggle with long-horizon planning due to the transient nature of purely prompting-based reasoning. While existing skill induction methods mitigate this by distilling experience into state-blind parameterized scripts, they fail to capture the conditional logic required for robust execution in dynamic environments. In this paper, we propose Neuro-Symbolic Skill Induction (NSI), a framework that lifts interaction traces into modular, \textit{logic-grounded} programs. By synthesizing explicit control flows and dynamic variable binding, NSI empowers agents to discover \textit{when} and \textit{why} to act. This paradigm enables the efficient generalization, allowing agents to induce skills from few-shot examples and flexibly adapt to unseen goals. Experiments on a series of agentic tasks demonstrate that NSI consistently outperforms state-of-the-art baselines, empowering agents to self-evolve into architects of logic-grounded skills.
title Lifting Traces to Logic: Programmatic Skill Induction with Neuro-Symbolic Learning for Long-Horizon Agentic Tasks
topic Artificial Intelligence
url https://arxiv.org/abs/2605.01293