# LEMON: Local Explanations via Modality-aware OptimizatioN

Source: [LEMON: Local Explanations via Modality-aware OptimizatioN](<https://www.sota2.com/research/paper/6982d287a8b01e76cc8e92f1>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-05T05:00:04.050Z

Authors: Yu Qin, Phillip Sloan, Raul Santos-Rodriguez, Majid Mirmehdi, Telmo de Menezes e Silva Filho

Published: 2026-02-02T20:40:59.000Z

## Abstract

Multimodal models are ubiquitous, yet existing explainability methods are often single-modal, architecture-dependent, or too computationally expensive to run at scale. We introduce LEMON (Local Explanations via Modality-aware OptimizatioN), a model-agnostic framework for local explanations of multimodal predictions. LEMON fits a single modality-aware surrogate with group-structured sparsity to produce unified explanations that disentangle modality-level contributions and feature-level attributions. The approach treats the predictor as a black box and is computationally efficient, requiring relatively few forward passes while remaining faithful under repeated perturbations. We evaluate LEMON on vision-language question answering and a clinical prediction task with image, text, and tabular inputs, comparing against representative multimodal baselines. Across backbones, LEMON achieves competitive deletion-based faithfulness while reducing black-box evaluations by 35-67 times and runtime by 2-8 times compared to strong multimodal baselines.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2602.02786>)

[PDF](<https://arxiv.org/pdf/2602.02786.pdf>)
