{"url":"/sota/conversational-response-selection-on-ubuntu-1","task":{"name":"Conversational Response Selection","url":"/task/conversational-response-selection","note":null},"dataset":{"name":"Ubuntu Dialogue (v1, Ranking)","url":"/dataset/ubuntu-dialogue-corpus"},"category":"Natural Language Processing","categories":["Natural Language Processing"],"category_note":null,"description":"Conversational response selection refers to the task of identifying the most relevant response to a given input sentence from a collection of sentences.","description_from":"task","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","rank":"the archive's row order at snapshot; not re-ranked","rows_end_at":"2025-07-28","rows_withheld_as_spam":0,"metric_values":"the archive's strings, untouched"},"metrics":["R10@1","R10@2","R10@5","R2@1"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"R10@1":null,"R10@2":null,"R10@5":null,"R2@1":null}},"counts":{"rows":25,"rows_with_code":18,"rows_with_paper_page":25,"rows_dated":25,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"Dial-MAE","metrics":{"R10@1":"0.918","R10@2":"0.964","R10@5":"0.993"},"uses_additional_data":false,"paper_date":"2023-06-07","paper":"/paper/contextual-masked-auto-encoder-for-retrieval","paper_url":"https://arxiv.org/abs/2306.04357v5","paper_title":"Dial-MAE: ConTextual Masked Auto-Encoder for Retrieval-based Dialogue Systems","code":"https://github.com/suu990901/Dial-MAE","n_code_links":1,"syntology":null},{"rank_in_archive_order":2,"model":"BERT-FP+EDHNS","metrics":{"R10@1":"0.917","R10@2":"0.965","R10@5":"0.994"},"uses_additional_data":false,"paper_date":"2024-08-16","paper":"/paper/efficient-dynamic-hard-negative-sampling-for","paper_url":"https://aclanthology.org/2024.nlp4convai-1.6","paper_title":"Efficient Dynamic Hard Negative Sampling for Dialogue Selection","code":"https://github.com/hanjanghoon/EDHNS","n_code_links":1,"syntology":null},{"rank_in_archive_order":3,"model":"Uni-Enc+BERT-FP","metrics":{"R10@1":"0.916","R10@2":"0.965","R10@5":"0.994"},"uses_additional_data":false,"paper_date":"2021-06-02","paper":"/paper/global-selector-a-new-benchmark-dataset-and","paper_url":"https://arxiv.org/abs/2106.01263v5","paper_title":"Uni-Encoder: A Fast and Accurate Response Selection Paradigm for Generation-Based Dialogue Systems","code":"https://github.com/dll-wu/uni-encoder","n_code_links":1,"syntology":null},{"rank_in_archive_order":4,"model":"BERT-FP","metrics":{"R10@1":"0.911","R10@2":"0.962","R10@5":"0.994"},"uses_additional_data":false,"paper_date":"2021-05-24","paper":"/paper/fine-grained-post-training-for-improving","paper_url":"https://aclanthology.org/2021.naacl-main.122","paper_title":"Fine-grained Post-training for Improving Retrieval-based Dialogue Systems","code":"https://github.com/hanjanghoon/BERT_FP","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"BERT-UMS+FGC","metrics":{"R10@1":"0.886","R10@2":"0.948","R10@5":"0.990"},"uses_additional_data":false,"paper_date":"2021-11-19","paper":"/paper/small-changes-make-big-differences-improving","paper_url":"https://arxiv.org/abs/2111.10154v2","paper_title":"Small Changes Make Big Differences: Improving Multi-turn Response Selection in Dialogue Systems via Fine-Grained Contrastive Learning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"Uni-Encoder","metrics":{"R10@1":"0.886","R10@2":"0.946","R10@5":"0.989"},"uses_additional_data":false,"paper_date":"2021-06-02","paper":"/paper/global-selector-a-new-benchmark-dataset-and","paper_url":"https://arxiv.org/abs/2106.01263v5","paper_title":"Uni-Encoder: A Fast and Accurate Response Selection Paradigm for Generation-Based Dialogue Systems","code":"https://github.com/dll-wu/uni-encoder","n_code_links":1,"syntology":null},{"rank_in_archive_order":7,"model":"BERT-SL","metrics":{"R10@1":"0.884","R10@2":"0.946","R10@5":"0.990","R2@1":"0.975"},"uses_additional_data":false,"paper_date":"2020-09-14","paper":"/paper/learning-an-effective-context-response","paper_url":"https://arxiv.org/abs/2009.06265v1","paper_title":"Learning an Effective Context-Response Matching Model with Self-Supervised Tasks for Retrieval-based Dialogues","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":8,"model":"Poly-encoder","metrics":{"R10@1":"0.882","R10@2":"0.949","R10@5":"0.990"},"uses_additional_data":false,"paper_date":"2019-04-22","paper":"/paper/190501969","paper_url":"https://arxiv.org/abs/1905.01969v4","paper_title":"Poly-encoders: Transformer Architectures and Pre-training Strategies for Fast and Accurate Multi-sentence Scoring","code":"https://github.com/sfzhou5678/PolyEncoder","n_code_links":7,"syntology":null},{"rank_in_archive_order":9,"model":"UMS_BERT+","metrics":{"R10@1":"0.875","R10@2":"0.942","R10@5":"0.988"},"uses_additional_data":false,"paper_date":"2020-09-10","paper":"/paper/do-response-selection-models-really-know-what","paper_url":"https://arxiv.org/abs/2009.04703v2","paper_title":"Do Response Selection Models Really Know What's Next? Utterance Manipulation Strategies for Multi-turn Response Selection","code":"https://github.com/taesunwhang/UMS-ResSel","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"BERT-VFT","metrics":{"R10@1":"0.855","R10@2":"0.928","R10@5":"0.985"},"uses_additional_data":false,"paper_date":"2019-08-13","paper":"/paper/domain-adaptive-training-bert-for-response","paper_url":"https://arxiv.org/abs/1908.04812v2","paper_title":"An Effective Domain Adaptive Post-Training Method for BERT in Response Selection","code":"https://github.com/taesunwhang/BERT-ResSel","n_code_links":1,"syntology":null},{"rank_in_archive_order":11,"model":"SA-BERT","metrics":{"R10@1":"0.855","R10@2":"0.928","R10@5":"0.983","R2@1":"0.965"},"uses_additional_data":false,"paper_date":"2020-04-07","paper":"/paper/speaker-aware-bert-for-multi-turn-response","paper_url":"https://arxiv.org/abs/2004.03588v2","paper_title":"Speaker-Aware BERT for Multi-Turn Response Selection in Retrieval-Based Chatbots","code":"https://github.com/JasonForJoy/BERT-for-Response-Selection","n_code_links":2,"syntology":null},{"rank_in_archive_order":12,"model":"WDMN","metrics":{"R10@1":"0.821","R10@2":"0.911","R10@5":"0.981","R2@1":"0.957"},"uses_additional_data":false,"paper_date":"2021-08-17","paper":"/paper/response-ranking-with-multi-types-of-deep","paper_url":"https://dl.acm.org/doi/abs/10.1145/3462207","paper_title":"Response Ranking with Multi-types of Deep Interactive Representations in Retrieval-based Dialogues","code":"https://github.com/RayXu14/WDMN","n_code_links":1,"syntology":null},{"rank_in_archive_order":13,"model":"MSN","metrics":{"R10@1":"0.800","R10@2":"0.899","R10@5":"0.978"},"uses_additional_data":false,"paper_date":"2019-11-01","paper":"/paper/multi-hop-selector-network-for-multi-turn","paper_url":"https://aclanthology.org/D19-1011","paper_title":"Multi-hop Selector Network for Multi-turn Response Selection in Retrieval-based Chatbots","code":"https://github.com/chunyuanY/Dialogue","n_code_links":1,"syntology":null},{"rank_in_archive_order":14,"model":"ESIM","metrics":{"R10@1":"0.796","R10@2":"0.894","R10@5":"0.975"},"uses_additional_data":false,"paper_date":"2019-01-09","paper":"/paper/sequential-attention-based-network-for-noetic","paper_url":"https://arxiv.org/abs/1901.02609v3","paper_title":"Sequential Attention-based Network for Noetic End-to-End Response Selection","code":"https://github.com/alibaba/esim-response-selection","n_code_links":4,"syntology":null},{"rank_in_archive_order":15,"model":"IoI-local","metrics":{"R10@1":"0.796","R10@2":"0.894","R10@5":"0.974","R2@1":"0.947"},"uses_additional_data":false,"paper_date":"2019-07-01","paper":"/paper/one-time-of-interaction-may-not-be-enough-go","paper_url":"https://aclanthology.org/P19-1001","paper_title":"One Time of Interaction May Not Be Enough: Go Deep with an Interaction-over-Interaction Network for Response Selection in Dialogues","code":"https://github.com/chongyangtao/IOI","n_code_links":1,"syntology":null},{"rank_in_archive_order":16,"model":"IMN","metrics":{"R10@1":"0.794","R10@2":"0.889","R10@5":"0.974","R2@1":"0.946"},"uses_additional_data":false,"paper_date":"2019-01-07","paper":"/paper/interactive-matching-network-for-multi-turn","paper_url":"https://arxiv.org/abs/1901.01824v2","paper_title":"Interactive Matching Network for Multi-Turn Response Selection in Retrieval-Based Chatbots","code":"https://github.com/JasonForJoy/IMN","n_code_links":1,"syntology":null},{"rank_in_archive_order":17,"model":"TripleNet","metrics":{"R10@1":"0.790","R10@2":"0.885","R10@5":"0.970","R2@1":"0.943"},"uses_additional_data":false,"paper_date":"2019-09-24","paper":"/paper/triplenet-triple-attention-network-for-multi","paper_url":"https://arxiv.org/abs/1909.10666v2","paper_title":"TripleNet: Triple Attention Network for Multi-Turn Response Selection in Retrieval-based Chatbots","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":18,"model":"DAM-Semi","metrics":{"R10@1":"0.785","R10@2":"0.883","R10@5":"0.974","R2@1":"0.944"},"uses_additional_data":false,"paper_date":"2019-11-01","paper":"/paper/sampling-matters-an-empirical-study-of","paper_url":"https://aclanthology.org/D19-1128","paper_title":"Sampling Matters! An Empirical Study of Negative Sampling Strategies for Learning of Matching Models in Retrieval-based Dialogue Systems","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":19,"model":"DAM","metrics":{"R10@1":"0.767","R10@2":"0.874","R10@5":"0.969","R2@1":"0.938"},"uses_additional_data":false,"paper_date":"2018-07-01","paper":"/paper/multi-turn-response-selection-for-chatbots","paper_url":"https://aclanthology.org/P18-1103","paper_title":"Multi-Turn Response Selection for Chatbots with Deep Attention Matching Network","code":"https://github.com/baidu/Dialogue","n_code_links":1,"syntology":null},{"rank_in_archive_order":20,"model":"DAM-MG","metrics":{"R10@1":"0.753","R2@1":"0.935"},"uses_additional_data":false,"paper_date":"2019-08-26","paper":"/paper/multi-granularity-representations-of-dialog","paper_url":"https://arxiv.org/abs/1908.09890v1","paper_title":"Multi-Granularity Representations of Dialog","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":21,"model":"DUA","metrics":{"R10@1":"0.752","R10@2":"0.868","R10@5":"0.962"},"uses_additional_data":false,"paper_date":"2018-06-24","paper":"/paper/modeling-multi-turn-conversation-with-deep","paper_url":"http://arxiv.org/abs/1806.09102v2","paper_title":"Modeling Multi-turn Conversation with Deep Utterance Aggregation","code":"https://github.com/cooelf/DeepUtteranceAggregation","n_code_links":1,"syntology":null},{"rank_in_archive_order":22,"model":"SMN","metrics":{"R10@1":"0.726","R10@2":"0.822","R10@5":"0.960","R2@1":"0.926"},"uses_additional_data":false,"paper_date":"2016-12-06","paper":"/paper/sequential-matching-network-a-new","paper_url":"http://arxiv.org/abs/1612.01627v2","paper_title":"Sequential Matching Network: A New Architecture for Multi-turn Response Selection in Retrieval-based Chatbots","code":"https://github.com/MarkWuNLP/MultiTurnResponseSelection","n_code_links":3,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":23,"model":"Multi-View","metrics":{"R10@1":"0.662","R10@2":"0.801","R10@5":"0.951","R2@1":"0.908"},"uses_additional_data":false,"paper_date":"2016-11-01","paper":"/paper/multi-view-response-selection-for-human","paper_url":"https://aclanthology.org/D16-1036","paper_title":"Multi-view Response Selection for Human-Computer Conversation","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":24,"model":"Dual-BiLSTM","metrics":{"R10@1":"0.630","R10@2":"0.780","R10@5":"0.944","R2@1":"0.895"},"uses_additional_data":false,"paper_date":"2015-10-13","paper":"/paper/improved-deep-learning-baselines-for-ubuntu","paper_url":"http://arxiv.org/abs/1510.03753v2","paper_title":"Improved Deep Learning Baselines for Ubuntu Corpus Dialogs","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":25,"model":"Dual-LSTM","metrics":{"R10@1":"0.604","R10@2":"0.745","R10@5":"0.926","R2@1":"0.878"},"uses_additional_data":false,"paper_date":"2015-06-30","paper":"/paper/the-ubuntu-dialogue-corpus-a-large-dataset-1","paper_url":"http://arxiv.org/abs/1506.08909v3","paper_title":"The Ubuntu Dialogue Corpus: A Large Dataset for Research in Unstructured Multi-Turn Dialogue Systems","code":"https://github.com/facebookresearch/ParlAI","n_code_links":21,"syntology":{"n_ran":1,"n_unverified":25,"n_samples":26,"n_pointer_only_licence":1}}],"since_archive":{"claim":"Results that newer papers report for their own method, placed here by Syntology. A model pointed at the cell in the paper's own table; the number was read from that cell and checked against this leaderboard's metric, dataset, split and scale; an independent check that saw this leaderboard's other rows and every other leaderboard on the same dataset accepted it. Not reviewed by the paper's authors or by the archive's editors, and not ranked against the archive rows.","extraction_file_present":true,"measurement":{"test_papers":883,"papers_with_output":881,"judged_true":108,"judged":110,"wilson95_lower":0.9361,"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication of accepted entries on a held-out split of archive papers, rules frozen before the test","coverage":{"sentence":"Syntology has checked 6,264 of the 9,581 papers on this site that are newer than the archive; results from the others appear after they are checked.","complete":false,"papers_newer_than_archive":9581,"papers_checked":6264,"papers_extracted_not_yet_verified":0,"boards_without_verdict":2,"papers_not_yet_extracted":3316},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-09-24T18:15:14+00:00","claim":"Per row: N of M harvested code samples from that row's paper executed on a synthesized fixture; the other M-N are unverified. Not a reproduction of the row's number; not a correctness claim. n_pointer_only_licence counts samples the site points at rather than redistributes (a licence axis, independent of ran/unverified).","rows_with_graph_line":2,"rows_with_any_sample_ran":1,"distinct_papers_with_graph_line":2,"distinct_papers_with_any_sample_ran":1,"samples_over_distinct_papers":{"n_ran":1,"n_unverified":28,"n_samples":29,"n_pointer_only_licence":1,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":1,"n_unverified":28,"n_samples":29,"n_pointer_only_licence":1,"note":"row-weighted: a paper behind several rows is counted once per row; inflated relative to samples_over_distinct_papers by design, kept for readers summing the per-row syntology blocks"}}}