Lifting Traces to Logic: Programmatic Skill Induction with Neuro-Symbolic Learning for Long-Horizon Agentic Tasks

TL;DR

This paper introduces Neuro-Symbolic Skill Induction (NSI), a framework that converts interaction traces into logic-grounded programs, enhancing agent reasoning, generalization, and adaptability in long-horizon tasks.

Contribution

NSI is the first method to synthesize explicit control flows and variable bindings from interaction traces, enabling flexible, logic-based skill induction from few-shot examples.

Findings

NSI outperforms state-of-the-art baselines on agentic tasks.

Agents can induce skills from few-shot examples.

NSI enables agents to adapt to unseen goals.

Abstract

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.