Year: 2026
Type: conference-paper
Abstract: Large Multimodal Models (LMMs) achieve state-of-the-art performance in high-stakes domains like healthcare, yet their reasoning remains opaque.Attention and saliency-based methods often fail to faithfully represent the model's decision process, particularly when integrating heterogeneous modalities.We introduce Tree-of-Evidence (ToE), an inferencetime search algorithm that frames interpretability as a discrete optimization problem.Rather than relying on soft attention weights, ToE employs lightweight Evidence Bottlenecks that score coarse units of data (e.g., vital-sign windows, report chunks) and performs a beam search to identify the compact evidence set required to reproduce the model's prediction.We evaluate ToE across six tasks spanning three datasets and two domains, including clinical prediction on MIMIC-IV, cross-center validation on eICU, and non-clinical fault detection on LEMMA-RCA.ToE retains over 98% of full-model AUROC with as few as five evidence units, achieves higher decision agreement and lower fidelity error than LIME, SHAP, saliency, and concept-bottleneck baselines under sparse budgets, and outperforms LLMs up to 70B parameters.ToE therefore provides a practical mechanism for auditing multimodal models.Phase I: Time-Series (TS) and Notes classifiers training Phase II: Selector Training Phase III: Tree of Evidence (Inference Search) … 𝑾
Institution: Georgia Institute of Technology
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7166806271