CircuitSynth: Reliable Synthetic Data Generation

arXiv:2604.10114v1 Announce Type: cross Abstract: The generation of high-fidelity synthetic data is a cornerstone of modern machine learning, yet Large Language Models (LLMs) frequently suffer from hallucinations, logical inconsistencies, and mode collapse when tasked with structured generation. Existing approaches, such as prompting or retrieval-augmented generation, lack the mechanisms to balance linguistic expressivity with formal guarantees regarding validity and coverage. To address this, we propose CircuitSynth, a novel neuro-symbolic framework that decouples semantic reasoning from surface realization. By distilling the reasoning capabilities of a Teacher LLM into a Probabilistic Sentential Decision Diagram (PSDD), CircuitSynth creates a tractable semantic prior that structurally enforces hard logical constraints. Furthermore, we introduce a convex optimization mechanism to rigorously satisfy soft distributional goals. Empirical evaluations across diverse benchmarks demonstrate that CircuitSynth achieves 100% Schema Validity even in complex logic puzzles where unconstrained baselines fail (12.4%) while significantly outperforming state-of-the-art methods in rare-combination coverage.

Related

• Private Seeds, Public LLMs: Realistic and Privacy-Preserving Synthetic Data Generation
• Structured Exploration and Exploitation of Label Functions for Automated Data Annotation
• Decoding by Perturbation: Mitigating MLLM Hallucinations via Dynamic Textual Perturbation
• Faithful-First Reasoning, Planning, and Acting for Multimodal LLMs
• ProofSketcher: Hybrid LLM + Lightweight Proof Checker for Reliable Math/Logic Reasoning

Source: arXiv cs.AI | 2026-04-14