Search papers, labs, and topics across Lattice.
This paper introduces DIFFIE, a novel approach to Open Information Extraction (OpenIE) that leverages conditional discrete diffusion as a mechanism for generating relational triplets. By employing independent reverse-diffusion trajectories over role tags, DIFFIE produces a diverse pool of candidate triplets that can be flexibly clustered and ranked, decoupling the extraction budget from training. The method achieves state-of-the-art performance on the CaRB benchmark and surpasses the leading rule-based system, demonstrating that diffusion stochasticity is a powerful tool for structured prediction tasks with multiple valid outputs.
DIFFIE reveals that leveraging diffusion stochasticity can significantly enhance the efficiency and effectiveness of OpenIE systems, outperforming traditional methods.
A single sentence often expresses multiple valid relational triplets, which makes Open Information Extraction (OpenIE) fundamentally a multi-output task. Existing neural systems handle this by autoregressive generation, which is flexible but slow and prone to redundancy, or by fixed-slot prediction, which is efficient but couples the extraction budget to training. We introduce DIFFIE which instead treats the stochasticity of conditional discrete diffusion as the extraction mechanism itself: independent reverse-diffusion trajectories over per-token role tags produce a pool of candidate triplets, which are clustered under lenient matching and ranked to form the output. Both the pool size and the number of returned extractions are inference-time choices, decoupling the extraction budget from training and exposing test-time compute as a tunable axis. DIFFIE achieves the new state of the art in CaRB (1-1) both F1 and AUC, and outperforms the strongest rule-based system (ClausIE) in BenchIE; it also remains competitive in standard CaRB and WiRe57 evaluations, giving the best average score among systems that report all four benchmarks. Ablations show that uniform discrete diffusion outperforms absorbing state diffusion in our setting, and that a matched non-diffusion stochastic tagger does not reproduce its gains. Our results indicate that diffusion stochasticity is an effective mechanism for structured prediction tasks with multiple valid outputs.
