{"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/dissecting-and-mitigating-diffusion-bias-via","title":"Dissecting and Mitigating Diffusion Bias via Mechanistic Interpretability","arxiv_id":"2503.20483","date":"2025-03-26","proceeding":"CVPR 2025 1","authors":["Yingdong Shi","Changming Li","Yifan Wang","Yongxiang Zhao","Anqi Pang","Sibei Yang","Jingyi Yu","Kan Ren"],"abstract":"Diffusion models have demonstrated impressive capabilities in synthesizing diverse content. However, despite their high-quality outputs, these models often perpetuate social biases, including those related to gender and race. These biases can potentially contribute to harmful real-world consequences, reinforcing stereotypes and exacerbating inequalities in various social contexts. While existing research on diffusion bias mitigation has predominantly focused on guiding content generation, it often neglects the intrinsic mechanisms within diffusion models that causally drive biased outputs. In this paper, we investigate the internal processes of diffusion models, identifying specific decision-making mechanisms, termed bias features, embedded within the model architecture. By directly manipulating these features, our method precisely isolates and adjusts the elements responsible for bias generation, permitting granular control over the bias levels in the generated content. Through experiments on both unconditional and conditional diffusion models across various social bias attributes, we demonstrate our method's efficacy in managing generation distribution while preserving image quality. We also dissect the discovered model mechanism, revealing different intrinsic features controlling fine-grained aspects of generation, boosting further research on mechanistic interpretability of diffusion models.","url_abs":"https://arxiv.org/abs/2503.20483v1","url_pdf":"https://arxiv.org/pdf/2503.20483v1.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":"dissecting-and-mitigating-diffusion-bias-via","repo_url":"https://github.com/foundation-model-research/DiffLens","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"age-unbiased","task_name":"Age/Unbiased"},{"task_slug":"decision-making","task_name":"Decision Making"},{"task_slug":"image-generation","task_name":"Image Generation"},{"task_slug":"interpretability-techniques-for-deep-learning","task_name":"Interpretability Techniques for Deep Learning"},{"task_slug":"race-unbiased","task_name":"Race/Unbiased"}],"methods":[{"method_slug":"diffusion","method_name":"Diffusion"},{"method_slug":"sparse-autoencoder","method_name":"Sparse Autoencoder"},{"method_slug":"k-sparse-autoencoder","method_name":"k-Sparse Autoencoder"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2503.20483","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2503.20483"}},"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":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/foundation-model-research/DiffLens","reach":null}],"summary":{"ran":1},"by_repo_kind":{"listed":{"samples":1,"ran":1,"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":1,"samples":[{"code_sha256_prefix":"38524b27f75946ea","entry":"FastAutoencoder","repo":"foundation-model-research/DiffLens","repo_kind":"listed","path":"difflens/locate/locate_main/ksae_model.py","file_url":"https://github.com/foundation-model-research/DiffLens/blob/HEAD/difflens/locate/locate_main/ksae_model.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"38524b27f75946ea"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}