KAPSO: A Knowledge-grounded framework for Autonomous Program Synthesis and Optimization

2026-01-29

AI Summary

This work proposes a modular autonomous programming framework to address key challenges in long-horizon coding tasks, including fragile debugging, frequent loss of experimental state, and insufficient reuse of domain knowledge. The framework embeds program synthesis within a long-term optimization loop structured around the “conceive–generate–execute–evaluate–learn” cycle. It innovatively integrates a Git-native experimentation engine, a structured multi-source knowledge system encompassing code, documentation, and research papers, and a cognitive memory layer grounded in execution logs. This integration enables end-to-end optimization that is reproducible, knowledge-driven, and capable of accumulating experiential learning. Experiments on MLE-Bench and ALE-Bench demonstrate that the approach significantly reduces recurrent errors, accelerates convergence, and enhances overall performance.

Technology Category

Natural Language Processing: Code Generation / Program Synthesis from Natural LanguageSearch and Optimization: Learning to SearchPlanning, Routing, and Scheduling: Learning for Planning and Scheduling

Application Category

Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for rankingEconomics, Online Markets and Human Computation: Cost models of using LLMs in production systems

Abstract

We introduce KAPSO, a modular framework for autonomous program synthesis and optimization. Given a natural language goal and an evaluation method, KAPSO iteratively performs ideation, code synthesis and editing, execution, evaluation, and learning to improve a runnable artifact toward measurable objectives. Rather than treating synthesis as the endpoint, KAPSO uses synthesis as an operator within a long-horizon optimization loop, where progress is defined by evaluator outcomes. KAPSO targets long-horizon failures common in coding agents, including lost experimental state, brittle debugging, and weak reuse of domain expertise, by integrating three tightly coupled components. First, a git-native experimentation engine isolates each attempt as a branch, producing reproducible artifacts and preserving provenance across iterations. Second, a knowledge system ingests heterogeneous sources, including repositories, internal playbooks, and curated external resources such as documentation, scientific papers, and web search results, and organizes them into a structured representation that supports retrieval over workflows, implementations, and environment constraints. Third, a cognitive memory layer coordinates retrieval and maintains an episodic store of reusable lessons distilled from experiment traces (run logs, diffs, and evaluator feedback), reducing repeated error modes and accelerating convergence. We evaluated KAPSO on MLE-Bench (Kaggle-style ML competitions) and ALE-Bench (AtCoder heuristic optimization), and report end-to-end performance. Code Available at: https://github.com/Leeroo-AI/kapso

Problem

• autonomous program synthesis
• long-horizon failures
• domain expertise reuse
• experimental state preservation
• brittle debugging

Innovation

• autonomous program synthesis
• knowledge-grounded optimization
• git-native experimentation
• cognitive memory
• structured knowledge retrieval