Steerable Instruction Following Coding Data Synthesis with Actor-Parametric Schema Co-Evolution

TL;DR

This paper introduces IFCodeEvolve, a novel framework for generating instruction-following coding data using actor-schema co-evolution, significantly improving model performance and providing a new benchmark.

Contribution

The paper presents a new actor-schema co-evolution approach for synthesizing instruction-paired coding data and introduces IFCodeBench, a verified benchmark for evaluation.

Findings

IFCodeEvolve boosts base model performance to SOTA levels.

The framework effectively navigates large instruction spaces with MCTS.

IFCodeEvolve's data synthesis improves instruction-following capabilities.

Abstract

Interpreting and following human instructions is a critical capability of large language models (LLMs) in automatic programming. However, synthesizing large-scale instruction-paired coding data remains largely unexplored and is particularly challenging when ensuring logical compatibility among multiple constraints. In this study, we propose IFCodeEvolve, an actor-schema co-evolution framework for instruction following coding data generation. By representing instructions as parametric function schema, we construct a library that covers the vast instruction space via dynamic constraint instantiation. Building upon this, Monte Carlo Tree Search (MCTS) sampler is applied to efficiently navigate this space, utilizing actor model feedback as a dynamic termination signal. Furthermore, to progressively explore challenging problems, we introduce a co-evolving paradigm that iteratively advances both the actor model and the schema library, via schema composition and mutation, based on sampler statistics. Empirical results demonstrate that IFCodeEvolve significantly boosts base model performance, with our 32B model achieving parity with proprietary SOTA models. Additionally, we contribute IFCodeBench, a comprehensive human-verified benchmark equipped with solutions and robust AST-based verification.