{"about":{"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.","site":"https://codewithpapers.app","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","syntology":{"site":"https://syntology.ai","developers":"https://syntology.ai/developers","mcp":{"server":"https://syntology.ai/mcp","transport":"streamable-http","server_card":"https://syntology.ai/.well-known/mcp/server-card.json","auth":{"type":"trial token, no account","trial_token":"https://syntology.ai/api/oauth/trial/token","method":"POST","docs":"https://syntology.ai/developers"}},"have":"https://syntology.ai/api/graph/have?x=<method, arXiv id or title> (free, answers coverage only)","paper_base":"https://syntology.ai/paper/","atlas_base":"https://app.syntology.ai/?focus="},"machine_readable":[{"url":"https://codewithpapers.app/llms.txt","what":"the machine catalog: every machine-readable file, counted"},{"url":"https://codewithpapers.app/index/manifest.json","what":"paper-to-code index by arXiv id, with Syntology's counts"},{"url":"https://codewithpapers.app/search/manifest.json","what":"site search index (titles, authors) and its files"},{"url":"https://codewithpapers.app/download","what":"bulk files: Syntology's layer, described there"},{"url":"https://codewithpapers.app/build_manifest.json","what":"the build record: inputs, counts, exclusions, probes"}]},"url":"/task/toxic-comment-classification","name":"Toxic Comment Classification","slug":"toxic-comment-classification","description_markdown":null,"categories":[{"name":"Natural Language Processing","url":"/area/natural-language-processing"}],"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","slug_source":"archive_url"},"counts":{"papers_tagged":28,"papers_with_code":13,"benchmarks":4,"benchmark_tables_in_archive":4,"benchmark_tables_shown":4,"benchmark_tables_withheld_as_spam":0,"benchmark_definition":"a leaderboard table with at least one row; benchmark_tables_shown also counts the zero-row tables; benchmark_tables_in_archive adds the tables withheld as spam","datasets":7,"subtasks":0,"parent_tasks":1},"benchmarks":[{"leaderboard":"/sota/toxic-comment-classification-on-civil","slug":"toxic-comment-classification-on-civil","dataset":"Civil Comments","dataset_url":"/dataset/civil-comments","rows_in_archive":22,"metrics":["GMB BPSN","GMB Subgroup","Micro F1","Precision","Recall","Macro F1","AUROC","GMB BNSP"],"first_row_in_archive_order":{"model":"RoBERTa Focal Loss","paper_title":"A benchmark for toxic comment classification on Civil Comments dataset","paper_url":"/paper/a-benchmark-for-toxic-comment-classification","paper_date":"2023-01-26","arxiv_id":"2301.11125","code_links":[{"title":"Nigiva/hatespeech-detection-models","url":"https://github.com/Nigiva/hatespeech-detection-models"}],"syntology":null}},{"leaderboard":"/sota/toxic-comment-classification-on-cad","slug":"toxic-comment-classification-on-cad","dataset":"CAD","dataset_url":"/dataset/cad","rows_in_archive":1,"metrics":["F1 on Toxic class"],"first_row_in_archive_order":{"model":"ContextRNN","paper_title":"Revisiting Contextual Toxicity Detection in Conversations","paper_url":"/paper/revisiting-contextual-toxicity-detection-in","paper_date":"2021-11-24","arxiv_id":"2111.12447","code_links":[],"syntology":null}},{"leaderboard":"/sota/toxic-comment-classification-on-germeval-2021-1","slug":"toxic-comment-classification-on-germeval-2021-1","dataset":"GermEval 2021 - Toxic Comments test set","dataset_url":null,"rows_in_archive":1,"metrics":["F1"],"first_row_in_archive_order":{"model":"GBERT/GELECTRA Ensemble","paper_title":"FHAC at GermEval 2021: Identifying German toxic, engaging, and fact-claiming comments with ensemble learning","paper_url":"/paper/fhac-at-germeval-2021-identifying-german","paper_date":"2021-09-07","arxiv_id":"2109.03094","code_links":[{"title":"dslaborg/germeval2021","url":"https://github.com/dslaborg/germeval2021"}],"syntology":null}},{"leaderboard":"/sota/toxic-comment-classification-on-jigsaw-toxic","slug":"toxic-comment-classification-on-jigsaw-toxic","dataset":"Jigsaw Toxic Comment Classification Dataset","dataset_url":"/dataset/toxic-comment-classification-challenge","rows_in_archive":1,"metrics":["Validation Accuracy"],"first_row_in_archive_order":{"model":"CapsNet","paper_title":"Evaluating The Effectiveness of Capsule Neural Network in Toxic Comment Classification using Pre-trained BERT Embeddings","paper_url":"/paper/evaluating-the-effectiveness-of-capsule","paper_date":"2023-10-12","arxiv_id":null,"code_links":[{"title":"TashinAhmed/HATE","url":"https://github.com/TashinAhmed/HATE"}],"syntology":null}}],"datasets":[{"url":"/dataset/civil-comments","name":"Civil Comments","full_name":"Jigsaw Unintended Bias in Toxicity Classification","num_papers_in_archive":156},{"url":"/dataset/cad","name":"CAD","full_name":"Contextual Abuse Dataset","num_papers_in_archive":8},{"url":"/dataset/uit-victsd","name":"UIT-ViCTSD","full_name":"UIT Vietnamese Constructive and Toxic Speech Detection","num_papers_in_archive":7},{"url":"/dataset/wikidetox","name":"WikiDetox","full_name":"Wikipedia Detox","num_papers_in_archive":7},{"url":"/dataset/toxic-comment-classification-challenge","name":"Jigsaw Toxic Comment Classification Dataset","full_name":"","num_papers_in_archive":4},{"url":"/dataset/fairprism","name":"FairPrism","full_name":"","num_papers_in_archive":2},{"url":"/dataset/korean-multi-label-hate-speech-dataset","name":"K-MHaS: Korean Multi-label Hate Speech Dataset","full_name":"","num_papers_in_archive":1}],"subtasks":[],"parent_tasks":[{"url":"/task/text-classification","name":"Text Classification"}],"papers":{"order":"repositories listed in the archive (most first), then date (newest first), then slug; the archive holds no stars","population":"every paper the archive tags with this task (distinct papers; spam titles not listed). 'of' counts the papers listed; until 2026-09-25 it counted only the papers with a code link, and that number is of_with_code","page":1,"pages_in_order":1,"rows_per_page":100,"shown":28,"of":28,"of_with_code":13,"tagged_in_all":28,"counts":{"archive_papers_tagged":28,"with_a_code_link":13,"where_syntology_ran_a_sample":3,"not_listed_spam_title":0,"listed":28,"listed_where_code_ran":3,"where_syntology_ran_a_sample_split":{"with_a_run_with_no_instrument_failure":1,"every_run_a_failure_of_syntologys_instrument":2,"listed_with_a_run_with_no_instrument_failure":1,"listed_every_run_a_failure_of_syntologys_instrument":2,"filter":{"states":["a run with no instrument failure","any run, instrument failures included"],"default":"a run with no instrument failure","note":"on the 'only where code ran' pages the default hides, in the browser, the rows where every run was a failure of Syntology's instrument; the second state shows them again. Rows are hidden, never re-ordered; these twins list every row"}},"definition":"distinct papers the archive tags; 'where Syntology ran a sample' counts papers with at least one harvested sample that ran, which is not a correctness claim"},"items":[{"url":"/paper/nlpguard-a-framework-for-mitigating-the-use","slug":"nlpguard-a-framework-for-mitigating-the-use","title":"NLPGuard: A Framework for Mitigating the Use of Protected Attributes by NLP Classifiers","date":"2024-07-01","arxiv_id":"2407.01697","repositories_listed":1,"syntology":null},{"url":"/paper/pytorch-frame-a-modular-framework-for-multi","slug":"pytorch-frame-a-modular-framework-for-multi","title":"PyTorch Frame: A Modular Framework for Multi-Modal Tabular Learning","date":"2024-03-31","arxiv_id":"2404.00776","repositories_listed":1,"syntology":{"n":2,"n_ran":2,"n_constructed":0,"n_ran_checked":1,"n_instrument":1,"n_unverified":0,"n_honours":0,"n_violates":0,"n_no_contract":1,"n_pointer_only":0,"phrase":"2 ran (of which 0 constructed an object rather than computing a result; 1 with no instrument failure: 0 honoured, 0 violated, 1 with no contract checked; 1 where Syntology's instrument failed) · 0 unverified","sample_list":"/paper/pytorch-frame-a-modular-framework-for-multi#ran","syntology_url":"https://syntology.ai/paper/2404.00776","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2404.00776"}},"official":{"repos":["pyg-team/pytorch-frame"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["official"]}}},{"url":"/paper/evaluating-the-effectiveness-of-capsule","slug":"evaluating-the-effectiveness-of-capsule","title":"Evaluating The Effectiveness of Capsule Neural Network in Toxic Comment Classification using Pre-trained BERT Embeddings","date":"2023-10-12","arxiv_id":null,"repositories_listed":1,"syntology":null},{"url":"/paper/critic-large-language-models-can-self-correct","slug":"critic-large-language-models-can-self-correct","title":"CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing","date":"2023-05-19","arxiv_id":"2305.11738","repositories_listed":1,"syntology":{"n":5,"n_ran":4,"n_constructed":0,"n_ran_checked":0,"n_instrument":4,"n_unverified":1,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":0,"phrase":"4 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 4 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/critic-large-language-models-can-self-correct#ran","syntology_url":"https://syntology.ai/paper/2305.11738","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2305.11738"}},"official":{"repos":["microsoft/ProphetNet"],"state":"official (archive's flag): 4 ran","n_ran":4,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":1,"ran_from_kinds":["official"]}}},{"url":"/paper/a-benchmark-for-toxic-comment-classification","slug":"a-benchmark-for-toxic-comment-classification","title":"A benchmark for toxic comment classification on Civil Comments dataset","date":"2023-01-26","arxiv_id":"2301.11125","repositories_listed":1,"syntology":null},{"url":"/paper/fhac-at-germeval-2021-identifying-german","slug":"fhac-at-germeval-2021-identifying-german","title":"FHAC at GermEval 2021: Identifying German toxic, engaging, and fact-claiming comments with ensemble learning","date":"2021-09-07","arxiv_id":"2109.03094","repositories_listed":1,"syntology":null},{"url":"/paper/data-integration-for-toxic-comment","slug":"data-integration-for-toxic-comment","title":"Data Integration for Toxic Comment Classification: Making More Than 40 Datasets Easily Accessible in One Unified Format","date":"2021-08-01","arxiv_id":null,"repositories_listed":1,"syntology":null},{"url":"/paper/from-hero-to-z-eroe-a-benchmark-of-low-level","slug":"from-hero-to-z-eroe-a-benchmark-of-low-level","title":"From Hero to Z\\'eroe: A Benchmark of Low-Level Adversarial Attacks","date":"2020-12-01","arxiv_id":null,"repositories_listed":1,"syntology":null},{"url":"/paper/from-hero-to-zeroe-a-benchmark-of-low-level","slug":"from-hero-to-zeroe-a-benchmark-of-low-level","title":"From Hero to Zéroe: A Benchmark of Low-Level Adversarial Attacks","date":"2020-10-12","arxiv_id":"2010.05648","repositories_listed":1,"syntology":{"n":3,"n_ran":2,"n_constructed":0,"n_ran_checked":0,"n_instrument":2,"n_unverified":1,"n_honours":0,"n_violates":0,"n_no_contract":0,"n_pointer_only":0,"phrase":"2 ran (of which 0 constructed an object rather than computing a result; 0 with no instrument failure: 0 honoured, 0 violated, 0 with no contract checked; 2 where Syntology's instrument failed) · 1 unverified","sample_list":"/paper/from-hero-to-zeroe-a-benchmark-of-low-level#ran","syntology_url":"https://syntology.ai/paper/2010.05648","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2010.05648"}},"official":{"repos":["yannikbenz/zeroe"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":1,"ran_from_kinds":["official"]}}},{"url":"/paper/trojaning-language-models-for-fun-and-profit","slug":"trojaning-language-models-for-fun-and-profit","title":"Trojaning Language Models for Fun and Profit","date":"2020-08-01","arxiv_id":"2008.00312","repositories_listed":1,"syntology":null},{"url":"/paper/can-we-achieve-more-with-less-exploring-data","slug":"can-we-achieve-more-with-less-exploring-data","title":"Can We Achieve More with Less? Exploring Data Augmentation for Toxic Comment Classification","date":"2020-07-02","arxiv_id":"2007.00875","repositories_listed":1,"syntology":null},{"url":"/paper/is-preprocessing-of-text-really-worth-your","slug":"is-preprocessing-of-text-really-worth-your","title":"Is preprocessing of text really worth your time for online comment classification?","date":"2018-06-07","arxiv_id":"1806.02908","repositories_listed":1,"syntology":null},{"url":"/paper/convolutional-neural-networks-for-toxic","slug":"convolutional-neural-networks-for-toxic","title":"Convolutional Neural Networks for Toxic Comment Classification","date":"2018-02-27","arxiv_id":"1802.09957","repositories_listed":1,"syntology":null},{"url":null,"slug":"applying-llms-to-active-learning-towards-cost","title":"Applying LLMs to Active Learning: Towards Cost-Efficient Cross-Task Text Classification without Manually Labeled Data","date":"2025-02-24","arxiv_id":"2502.16892","repositories_listed":0,"syntology":null},{"url":null,"slug":"beyond-toxic-toxicity-detection-datasets-are","title":"Beyond Toxic: Toxicity Detection Datasets are Not Enough for Brand Safety","date":"2023-03-27","arxiv_id":"2303.15110","repositories_listed":0,"syntology":null},{"url":null,"slug":"a-new-generation-of-perspective-api-efficient","title":"A New Generation of Perspective API: Efficient Multilingual Character-level Transformers","date":"2022-02-22","arxiv_id":"2202.11176","repositories_listed":0,"syntology":null},{"url":null,"slug":"a-survey-of-toxic-comment-classification","title":"A Survey of Toxic Comment Classification Methods","date":"2021-12-13","arxiv_id":"2112.06412","repositories_listed":0,"syntology":null},{"url":null,"slug":"identification-of-bias-against-people-with","title":"Identification of Bias Against People with Disabilities in Sentiment Analysis and Toxicity Detection Models","date":"2021-11-25","arxiv_id":"2111.13259","repositories_listed":0,"syntology":null},{"url":"/paper/revisiting-contextual-toxicity-detection-in","slug":"revisiting-contextual-toxicity-detection-in","title":"Revisiting Contextual Toxicity Detection in Conversations","date":"2021-11-24","arxiv_id":"2111.12447","repositories_listed":0,"syntology":null},{"url":null,"slug":"ss-bert-mitigating-identity-terms-bias-in","title":"SS-BERT: Mitigating Identity Terms Bias in Toxic Comment Classification by Utilising the Notion of \"Subjectivity\" and \"Identity Terms\"","date":"2021-09-06","arxiv_id":"2109.02691","repositories_listed":0,"syntology":null},{"url":null,"slug":"detox-at-germeval-2021-toxic-comment","title":"DeTox at GermEval 2021: Toxic Comment Classification","date":"2021-09-01","arxiv_id":null,"repositories_listed":0,"syntology":null},{"url":null,"slug":"ircologne-at-germeval-2021-toxicity","title":"IRCologne at GermEval 2021: Toxicity Classification","date":"2021-09-01","arxiv_id":null,"repositories_listed":0,"syntology":null},{"url":null,"slug":"universitat-regensburg-maxs-at-germeval-2021","title":"Universität Regensburg MaxS at GermEval 2021 Task 1: Synthetic Data in Toxic Comment Classification","date":"2021-09-01","arxiv_id":null,"repositories_listed":0,"syntology":null},{"url":null,"slug":"explaining-the-deep-natural-language","title":"Explaining the Deep Natural Language Processing by Mining Textual Interpretable Features","date":"2021-06-12","arxiv_id":"2106.06697","repositories_listed":0,"syntology":null},{"url":null,"slug":"vietnamese-open-domain-complaint-detection-in","title":"Vietnamese Complaint Detection on E-Commerce Websites","date":"2021-04-24","arxiv_id":"2104.11969","repositories_listed":0,"syntology":null},{"url":"/paper/constructive-and-toxic-speech-detection-for","slug":"constructive-and-toxic-speech-detection-for","title":"Constructive and Toxic Speech Detection for Open-domain Social Media Comments in Vietnamese","date":"2021-03-18","arxiv_id":"2103.10069","repositories_listed":0,"syntology":null},{"url":null,"slug":"detecting-aggression-and-toxicity-using-a","title":"Detecting Aggression and Toxicity using a Multi Dimension Capsule Network","date":"2019-08-01","arxiv_id":null,"repositories_listed":0,"syntology":null},{"url":null,"slug":"challenges-for-toxic-comment-classification","title":"Challenges for Toxic Comment Classification: An In-Depth Error Analysis","date":"2018-09-20","arxiv_id":"1809.07572","repositories_listed":0,"syntology":null}],"syntology_records":3,"syntology_note":"a paper without a record is not a recorded non-run: it may lack an arXiv id or simply be absent from the graph layer"},"key_notes":{"n_ran_checked":"legacy name, kept unchanged so existing readers do not break: it counts the samples that ran with no instrument failure (honoured, violated, and ran with no contract checked); it does not mean a contract was checked, and the pages print it as 'K with no instrument failure', not 'K checked'","n_constructed":"a sub-count of the samples that ran, never subtracted from them and never a failure: an executed sample whose run returned an instance of its own class (fixture_out_type equals the entry name): the run built an object and did not compute a result (Syntology's RAN record, counts.constructed)"},"since_archive_judged":{"papers":0,"papers_placed":0,"judged_by":{"model":"a model","models_as_recorded":[],"from":"lb_rows_run_verified.jsonl: each record's `model` field (the extractor's call that wrote decisions[])"},"note":"a model's judgement from the paper's own tables that they report on this leaderboard, not a result: no value from a paper that was not placed is printed or carried here; papers counts each paper once over these leaderboards","backfill":"8,886 of 18,637 newer papers checked so far (733 more had nothing to place; 41 could not be read)","backfill_complete":false,"leaderboards":[]},"papers_where_code_ran":3,"papers_where_code_ran_first_page":"/task/toxic-comment-classification/papers/ran/1","papers_where_code_ran_caption":"We ran code from the paper's repository; we did not run it on this task or check it against the task's benchmarks.","pages":[{"order":"ran","page":1,"of":1,"url":"/task/toxic-comment-classification/papers/ran/1","json":"/task/toxic-comment-classification/papers/ran/1.json","rows":3}],"description_links":{"kept":0,"unwrapped_to_text":0,"bare_urls_linked":0,"relative_images_dropped":0,"rule":"internal links are kept only when the target slug exists in the catalog"},"syntology":{"read_at":"2026-10-05T00:36:11+00:00","claim":"Per-sample execution status on synthesized fixtures ('ran N of M samples'); not a correctness claim and not a ranking signal.","status_vocabulary":{"ran_honours":"ran, honoured the contract we drafted","ran_violates":"ran, violated the contract we drafted","ran_draft_wrong":"ran; our contract draft was wrong, not the code","ran_fixture":"ran; the contract check could not run (RAISES, DEP_MISSING or TIMEOUT)","ran":"ran on a synthesized input","unverified":"unverified (harvested, no recorded run)"},"row_route":"each syntology object on a benchmark's first row and on a paper carries sample_list (this site's paper page, #ran), syntology_url (syntology.ai's page for the paper, only when that route probed live) and mcp (the agent call that lists the paper's harvested samples; not a verdict on the paper)","developers":"https://syntology.ai/developers"},"not_shown":{"libraries":"the archive has no per-task library table","trend_sparklines":"the Trend column of the benchmarks table was a rendered image; it is not in the archive","social_and_latest_sorts":"stars and social signals are not in the archive"},"record_sha256":"17d1aae1abe254ba914a61b44787cb400ad22efc42f4b70f624aecd33f29dc07","record_changed_at":"2026-10-05","record_changed_at_basis":"changed"}