Transductively Informed Inductive Program Synthesis

2025-05-20

AI Summary

In program synthesis, inductive and transductive approaches are typically integrated in isolation, lacking explicit modeling of their interplay—leading to limited generalization and accuracy. This paper proposes the first unified framework that explicitly models the interaction between inductive and transductive paradigms: an inductive model generates candidate programs, while a transductive model dynamically constrains and guides the search using input-output examples. The method integrates neural program synthesis, conditional sequence modeling, and example-driven transductive reasoning, incorporating joint training and search-space reweighting. Evaluated on string and list manipulation tasks, it achieves state-of-the-art performance with significantly improved solution rates. Generated programs exhibit higher syntactic correctness and semantic fidelity, especially under out-of-distribution (OOD) settings, where its generalization advantage is particularly pronounced.

Technology Category

• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Machine Learning: Neuro-Symbolic Learning
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Abstraction and reasoning in program synthesis has seen significant progress through both inductive and transductive paradigms. Inductive approaches generate a program or latent function from input-output examples, which can then be applied to new inputs. Transductive approaches directly predict output values for given inputs, effectively serving as the function themselves. Current approaches combine inductive and transductive models via isolated ensembling, but they do not explicitly model the interaction between both paradigms. In this work, we introduce acs{tiips}, a novel framework that unifies transductive and inductive strategies by explicitly modeling their interactions through a cooperative mechanism: an inductive model generates programs, while a transductive model constrains, guides, and refines the search to improve synthesis accuracy and generalization. We evaluate acs{tiips} on two widely studied program synthesis domains: string and list manipulation. Our results show that acs{tiips} solves more tasks and yields functions that more closely match optimal solutions in syntax and semantics, particularly in out-of-distribution settings, yielding state-of-the-art performance. We believe that explicitly modeling the synergy between inductive and transductive reasoning opens promising avenues for general-purpose program synthesis and broader applications.

Problem

• Unifies inductive and transductive program synthesis paradigms
• Improves synthesis accuracy and generalization via cooperative interaction
• Enhances performance in out-of-distribution program synthesis tasks

Innovation

• Combines inductive and transductive program synthesis
• Uses cooperative mechanism for interaction modeling
• Improves synthesis accuracy and generalization