{"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-2601-15115","title":"Training-Free and Interpretable Hateful Video Detection via Multi-stage Adversarial Reasoning","arxiv_id":"2601.15115","date":"2026-01-21","proceeding":null,"authors":["Shuonan Yang","Yuchen Zhang","Zeyu Fu"],"abstract":"Hateful videos pose serious risks by amplifying discrimination, inciting violence, and undermining online safety. Existing training-based hateful video detection methods are constrained by limited training data and lack of interpretability, while directly prompting large vision-language models often struggle to deliver reliable hate detection. To address these challenges, this paper introduces MARS, a training-free Multi-stage Adversarial ReaSoning framework that enables reliable and interpretable hateful content detection. MARS begins with the objective description of video content, establishing a neutral foundation for subsequent analysis. Building on this, it develops evidence-based reasoning that supports potential hateful interpretations, while in parallel incorporating counter-evidence reasoning to capture plausible non-hateful perspectives. Finally, these perspectives are synthesized into a conclusive and explainable decision. Extensive evaluation on two real-world datasets shows that MARS achieves up to 10% improvement under certain backbones and settings compared to other training-free approaches and outperforms state-of-the-art training-based methods on one dataset. In addition, MARS produces human-understandable justifications, thereby supporting compliance oversight and enhancing the transparency of content moderation workflows. The code is available at https://github.com/Multimodal-Intelligence-Lab-MIL/MARS.","url_abs":"https://arxiv.org/abs/2601.15115","url_pdf":"https://arxiv.org/pdf/2601.15115","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":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2601.15115","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2601.15115"}},"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/Multimodal-Intelligence-Lab-MIL/MARS","reach":null}],"summary":{"ran_draft_wrong":4,"ran_honours":1,"unverified":2},"by_repo_kind":{"found_in_text":{"samples":7,"ran":5,"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":0,"samples":[{"code_sha256_prefix":"f69eb6fd71b774b4","entry":"clean_json_response","repo":"Multimodal-Intelligence-Lab-MIL/MARS","repo_kind":"found_in_text","path":"code/MARS/gemini2.5flash.py","file_url":"https://github.com/Multimodal-Intelligence-Lab-MIL/MARS/blob/HEAD/code/MARS/gemini2.5flash.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f69eb6fd71b774b4"}},{"code_sha256_prefix":"e609dfa06255ac50","entry":"clean_text","repo":"Multimodal-Intelligence-Lab-MIL/MARS","repo_kind":"found_in_text","path":"code/MARS/gemini2.5flash.py","file_url":"https://github.com/Multimodal-Intelligence-Lab-MIL/MARS/blob/HEAD/code/MARS/gemini2.5flash.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e609dfa06255ac50"}},{"code_sha256_prefix":"6f8fcab1838b516d","entry":"create_turn_prompt","repo":"Multimodal-Intelligence-Lab-MIL/MARS","repo_kind":"found_in_text","path":"code/MARS/gemini2.5flash.py","file_url":"https://github.com/Multimodal-Intelligence-Lab-MIL/MARS/blob/HEAD/code/MARS/gemini2.5flash.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6f8fcab1838b516d"}},{"code_sha256_prefix":"11fb9487de1cfda6","entry":"get_frames_from_folder","repo":"Multimodal-Intelligence-Lab-MIL/MARS","repo_kind":"found_in_text","path":"code/MARS/gemini2.5flash.py","file_url":"https://github.com/Multimodal-Intelligence-Lab-MIL/MARS/blob/HEAD/code/MARS/gemini2.5flash.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"11fb9487de1cfda6"}},{"code_sha256_prefix":"578d1c267cb9b593","entry":"load_image_for_gemini","repo":"Multimodal-Intelligence-Lab-MIL/MARS","repo_kind":"found_in_text","path":"code/MARS/gemini2.5flash.py","file_url":"https://github.com/Multimodal-Intelligence-Lab-MIL/MARS/blob/HEAD/code/MARS/gemini2.5flash.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"578d1c267cb9b593"}},{"code_sha256_prefix":"82f67fdd61f3b486","entry":"classify_multimodal_content_single_turn","repo":"Multimodal-Intelligence-Lab-MIL/MARS","repo_kind":"found_in_text","path":"code/MARS/gemini2.5flash.py","file_url":"https://github.com/Multimodal-Intelligence-Lab-MIL/MARS/blob/HEAD/code/MARS/gemini2.5flash.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"82f67fdd61f3b486"}},{"code_sha256_prefix":"aadfaa0259ae42af","entry":"process_video_multi_turn","repo":"Multimodal-Intelligence-Lab-MIL/MARS","repo_kind":"found_in_text","path":"code/MARS/gemini2.5flash.py","file_url":"https://github.com/Multimodal-Intelligence-Lab-MIL/MARS/blob/HEAD/code/MARS/gemini2.5flash.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"aadfaa0259ae42af"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.CV","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}