{"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/better-explain-transformers-by-illuminating","title":"Better Explain Transformers by Illuminating Important Information","arxiv_id":"2401.09972","date":"2024-01-18","proceeding":null,"authors":["Linxin Song","Yan Cui","Ao Luo","Freddy Lecue","Irene Li"],"abstract":"Transformer-based models excel in various natural language processing (NLP) tasks, attracting countless efforts to explain their inner workings. Prior methods explain Transformers by focusing on the raw gradient and attention as token attribution scores, where non-relevant information is often considered during explanation computation, resulting in confusing results. In this work, we propose highlighting the important information and eliminating irrelevant information by a refined information flow on top of the layer-wise relevance propagation (LRP) method. Specifically, we consider identifying syntactic and positional heads as important attention heads and focus on the relevance obtained from these important heads. Experimental results demonstrate that irrelevant information does distort output attribution scores and then should be masked during explanation computation. Compared to eight baselines on both classification and question-answering datasets, our method consistently outperforms with over 3\\% to 33\\% improvement on explanation metrics, providing superior explanation performance. Our anonymous code repository is available at: https://github.com/LinxinS97/Mask-LRP","url_abs":"https://arxiv.org/abs/2401.09972v3","url_pdf":"https://arxiv.org/pdf/2401.09972v3.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"better-explain-transformers-by-illuminating","repo_url":"https://github.com/linxins97/mask-lrp","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"question-answering","task_name":"Question Answering"}],"methods":[{"method_slug":"focus","method_name":"Focus"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2401.09972","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2401.09972"}},"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":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/LinxinS97/Mask-LRP","reach":null}],"summary":{"unverified":3},"by_repo_kind":{"official":{"samples":3,"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":3,"samples":[{"code_sha256_prefix":"e119d216caac97a7","entry":"RelProp","repo":"LinxinS97/Mask-LRP","repo_kind":"official","path":"Transformer_Explanation/modules/layers_ours.py","file_url":"https://github.com/LinxinS97/Mask-LRP/blob/HEAD/Transformer_Explanation/modules/layers_ours.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":"e119d216caac97a7"}},{"code_sha256_prefix":"b35514ebfa211359","entry":"RelPropSimple","repo":"LinxinS97/Mask-LRP","repo_kind":"official","path":"Transformer_Explanation/modules/layers_ours.py","file_url":"https://github.com/LinxinS97/Mask-LRP/blob/HEAD/Transformer_Explanation/modules/layers_ours.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":"b35514ebfa211359"}},{"code_sha256_prefix":"4eb87f661550d443","entry":"forward_hook","repo":"LinxinS97/Mask-LRP","repo_kind":"official","path":"Transformer_Explanation/modules/layers_ours.py","file_url":"https://github.com/LinxinS97/Mask-LRP/blob/HEAD/Transformer_Explanation/modules/layers_ours.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":"4eb87f661550d443"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}