2026-10-12T11:21:38Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33814515 2026-09-15T17:58:47Z category_146 category_734 category_811 category_106 portal_5 item_type_3 month_year_09_2026

<p>Comparative summary of existing XAI methods and the proposed hybrid framework.</p> Arul Selvam P. (24950178) Tamije Selvy P. (24950181) Space Science Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Science Policy positioning hybrid xai outperform existing baselines deletion metrics across consistent evaluation protocol aligning methodological rigor additional computational overhead 15 &# 8211 unlike prior single intuitive reasoning paths aware design indicating hybrid explainability framework framework provides explanations benchmark datasets using benchmark datasets single techniques causal reasoning aware pipeline proposed framework trust ratings technically faithful systematically evaluated study advances specific corpora sentiment analysis results show representative architectures remain opaque regulatory demands question answering practical applications potential suitability paper introduces method pipeline method approaches main contributions llama ). limiting adoption introducing less intelligent systems integrates saliency explainable ai ensuring objective demand accountability comparable assessment centered visualization case studies bridging trust based methods based attribution achieve state <p>Comparative summary of existing XAI methods and the proposed hybrid framework.</p> 2026-09-15T17:58:37Z Dataset Dataset 10.1371/journal.pone.0343472.t001 https://figshare.com/articles/dataset/_p_Comparative_summary_of_existing_XAI_methods_and_the_proposed_hybrid_framework_p_/33814515 CC BY 4.0