DecomposeRL: Learning to Ask Useful, Informative, and Diverse Questions for Semi-Supervised, Traceable Claim Verification
For claim verification, DecomposeRL provides an accurate, traceable method that reduces the need for labeled data and large models.
DecomposeRL achieves 86.3 in-domain and 69.8 out-of-domain balanced accuracy across 11 claim-verification benchmarks, matching 32B models and GPT-4.1-mini while being 4x smaller, and outperforms baselines in semi-supervised settings with only 10% labeled data.
Claim verification splits between end-to-end classifiers that are accurate but yields no inspectable traces, and decomposition-based methods produce inspectable traces but lag performance on benchmark datasets. We propose DecomposeRL an accurate claim-verifier that produce inspectable traces. DecomposeRL frames decomposition as an RL policy trained with GRPO and a multi-faceted reward ensemble, enabling both fully supervised and semi-supervised learning from unlabeled claims. DecomposeRL addresses the prohibitive training cost of GRPO with a data-curation funnel that distills 115K fact-verification claims into a compact, learning-signal-dense subset of 5K claims. We show that a DecomposeRL-7B policy trained with full supervision on only ~5K curated claims achieves 86.3 in-domain and 69.8 out-of-domain balanced accuracy across 11 claim-verification benchmarks containing biomedical, political, scientific, and general-domain claims. Despite being 4x smaller, it matches 32B baselines and GPT-4.1-mini, and it further outperforms baselines in a semi-supervised setting with only 10% labeled claims data. Code, data, and models are available at https://dipta007.github.io/DecomposeRL