@inproceedings{niess-kern-2026-spandiffusion,
    title = "{S}pan{D}iffusion: Flow Matching over Continuous Span Masks for Financial Causal Question Answering",
    author = "Niess, Georg  and
      Kern, Roman",
    editor = "El-Haj, Mo  and
      Moreno Sandoval, Antonio  and
      Garcia-Serrano, Ana  and
      Chen, Chung-Chi  and
      Rayson, Paul  and
      Torterolo Orta, Yanco Amor  and
      Martinez, Paloma  and
      Porta, Jordi",
    booktitle = "The 7th Financial Narrative Processing Workshop",
    month = may,
    year = "2026",
    address = "Palma de Mallorca, Spain",
    publisher = "European Language Resources Association (ELRA)",
    url = "https://preview.aclanthology.org/revision-workflow/2026.fnp-1.15/",
    doi = "10.63317/4bm7v3ccgmjy",
    pages = "146--151",
    abstract = "We present SpanDiffusion, a continuous diffusion approach to extractive causal question answering for the FinCausal 2026 shared task. SpanDiffusion uses two Gaussian masks, continuous signals with peaks at the answer start and end positions, and learns to denoise them from pure noise through a dedicated transformer conditioned on frozen DeBERTa-v3-large embeddings with LoRA adapters (1.6M parameters). By replacing Denoising Diffusion Probabilistic Models (DDPM) with flow matching (rectified flow), we reduce denoising to only 20 Euler steps at inference. A systematic ablation across six diffusion variants and a span-classification baseline shows that LoRA adaptation is the dominant factor (+34 Exact Match points), followed by flow matching (+5.5 EM). However, the standard span classifier (85.8{\%} EM) outperforms our best diffusion model (83.0{\%} EM), suggesting that the denoiser does not yet justify its added complexity. We discuss tradeoffs between the interpretability of diffusion trajectories and classification accuracy."
}