{"ID":2829777,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2512.11412","arxiv_id":"2512.11412","title":"Task-Specific Sparse Feature Masks for Molecular Toxicity Prediction with Chemical Language Models","abstract":"Reliable in silico molecular toxicity prediction is a cornerstone of modern drug discovery, offering a scalable alternative to experimental screening. However, the black-box nature of state-of-the-art models remains a significant barrier to adoption, as high-stakes safety decisions demand verifiable structural insights alongside predictive performance. To address this, we propose a novel multi-task learning (MTL) framework designed to jointly enhance accuracy and interpretability. Our architecture integrates a shared chemical language model with task-specific attention modules. By imposing an L1 sparsity penalty on these modules, the framework is constrained to focus on a minimal set of salient molecular fragments for each distinct toxicity endpoint. The resulting framework is trained end-to-end and is readily adaptable to various transformer-based backbones. Evaluated on the ClinTox, SIDER, and Tox21 benchmark datasets, our approach consistently outperforms both single-task and standard MTL baselines. Crucially, the sparse attention weights provide chemically intuitive visualizations that reveal the specific fragments influencing predictions, thereby enhancing insight into the model's decision-making process.","short_abstract":"Reliable in silico molecular toxicity prediction is a cornerstone of modern drug discovery, offering a scalable alternative to experimental screening. However, the black-box nature of state-of-the-art models remains a significant barrier to adoption, as high-stakes safety decisions demand verifiable structural insights...","url_abs":"https://arxiv.org/abs/2512.11412","url_pdf":"https://arxiv.org/pdf/2512.11412v1","authors":"[\"Kwun Sy Lee\",\"Jiawei Chen\",\"Fuk Sheng Ford Chung\",\"Tianyu Zhao\",\"Zhenyuan Chen\",\"Debby D. Wang\"]","published":"2025-12-12T09:41:04Z","proceeding":"cs.CE","tasks":"[\"cs.CE\",\"cs.AI\",\"cs.CL\",\"cs.LG\",\"q-bio.BM\"]","methods":"[\"Transformer\",\"Language Model\"]","has_code":false}