{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/arxiv-2606-07533","title":"Bridging Traditional Explainability Methods and Multimodal Multilingual Models: An XAI-Based Analysis","arxiv_id":"2606.07533","date":"2026-04-21","proceeding":null,"authors":["Paweł Pozorski","Jakub Muszyński","Maria Ganzha"],"abstract":"Multimodal Large Language Models (MLLMs) effectively integrate text and audio to interpret context in complex interactive dialogues. However, the internal mechanisms by which heterogeneous modalities influence model behavior remain opaque. While Shapley Values (SV) provide a robust, model-agnostic framework for local explainability in text-based NLP, their extension to multimodal data is hindered by cross-channel dependencies, intricate dialogue structures, and the prohibitive computational complexity of dense audio representations. In this work, we formalize a multimodal extension of the Shapley Value framework, treating discrete text tokens and aligned audio segments as cooperative features. To ensure computational feasibility, we deploy a suite of efficient estimation strategies: exact SV computation for low-dimensional inputs and sampling-based approximations - including Monte Carlo permutations and stratified sampling with Neyman-optimal allocation - to minimize variance under constrained computational budgets. To resolve the granularity mismatch between modalities, we propose Spectrogram-Guided Phonetic Alignment (SGPA), a novel preprocessing method that maps high-frequency audio streams to interpretable, word-aligned segments. Our contribution is twofold: first, we provide an open-source, model-agnostic Python package and a companion GUI for the computation and interactive visualization of multimodal attributions. Second, we evaluate our framework using curated subsets of the VoiceBench and Infinity Instruct datasets across diverse multilingual scenarios. Our experimental results reveal that input modality is a primary driver of attribution volatility and demonstrate that standard syntactic importance proxies often fail to predict model attention in multimodal, cross-lingual contexts.","url_abs":"https://arxiv.org/abs/2606.07533","url_pdf":"https://arxiv.org/pdf/2606.07533","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2606.07533"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/mvishiu11/shap-mllm-explainer","reach":null}],"summary":{"unverified":4},"by_repo_kind":{"found_in_text":{"samples":4,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":4,"samples":[{"code_sha256_prefix":"185157c25050bcae","entry":"get_job","repo":"mvishiu11/shap-mllm-explainer","repo_kind":"found_in_text","path":"backend/app/services/progress.py","file_url":"https://github.com/mvishiu11/shap-mllm-explainer/blob/HEAD/backend/app/services/progress.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"185157c25050bcae"}},{"code_sha256_prefix":"1414be7289a441fb","entry":"get_job_logs","repo":"mvishiu11/shap-mllm-explainer","repo_kind":"found_in_text","path":"backend/app/services/job_logging.py","file_url":"https://github.com/mvishiu11/shap-mllm-explainer/blob/HEAD/backend/app/services/job_logging.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"1414be7289a441fb"}},{"code_sha256_prefix":"9985f23eff1df314","entry":"init_job","repo":"mvishiu11/shap-mllm-explainer","repo_kind":"found_in_text","path":"backend/app/services/progress.py","file_url":"https://github.com/mvishiu11/shap-mllm-explainer/blob/HEAD/backend/app/services/progress.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9985f23eff1df314"}},{"code_sha256_prefix":"3b6e8cf2c9fe5a9e","entry":"is_cancelled","repo":"mvishiu11/shap-mllm-explainer","repo_kind":"found_in_text","path":"backend/app/services/progress.py","file_url":"https://github.com/mvishiu11/shap-mllm-explainer/blob/HEAD/backend/app/services/progress.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"3b6e8cf2c9fe5a9e"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.CL","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}