{"url":"/sota/question-answering-on-babi","task":{"name":"Question Answering","url":"/task/question-answering","note":null},"dataset":{"name":"bAbi","url":null},"category":"Natural Language Processing","categories":["Miscellaneous","Natural Language Processing","Reasoning"],"category_note":null,"description":"Question answering can be segmented into domain-specific tasks like community question answering and knowledge-base question answering. Popular benchmark datasets for evaluation question answering systems include [SQuAD](/dataset/squad), [HotPotQA](/dataset/hotpotqa), [bAbI](/dataset/babi-1), [TriviaQA](/dataset/triviaqa), [WikiQA](/dataset/wikiqa), and many others. Models for question answering are typically evaluated on metrics like EM and F1. Some recent top performing models are T5 and XLNet.\r\n\r\n( Image credit: [SQuAD](https://rajpurkar.github.io/mlx/qa-and-squad/) )","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":["Accuracy (trained on 10k)","Accuracy (trained on 1k)","Mean Error Rate"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction); null = not inferred, chart draws points only","by_metric":{"Accuracy (trained on 10k)":"higher","Accuracy (trained on 1k)":"higher","Mean Error Rate":"lower"}},"counts":{"rows":14,"rows_with_code":10,"rows_with_paper_page":14,"rows_dated":14,"rows_using_additional_data":0},"rows":[{"rank_in_archive_order":1,"model":"STM","metrics":{"Accuracy (trained on 10k)":"99.85%","Mean Error Rate":"0.39%"},"uses_additional_data":false,"paper_date":"2020-02-10","paper":"/paper/self-assttentive-associative-memory","paper_url":"https://arxiv.org/abs/2002.03519v3","paper_title":"Self-Attentive Associative Memory","code":"https://github.com/thaihungle/SAM","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":3,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":2,"model":"QRN","metrics":{"Accuracy (trained on 10k)":"99.7%","Accuracy (trained on 1k)":"90.1%","Mean Error Rate":"0.3%"},"uses_additional_data":false,"paper_date":"2016-06-14","paper":"/paper/query-reduction-networks-for-question","paper_url":"http://arxiv.org/abs/1606.04582v6","paper_title":"Query-Reduction Networks for Question Answering","code":"https://github.com/voicy-ai/DialogStateTracking","n_code_links":2,"syntology":{"n_ran":0,"n_unverified":5,"n_samples":5,"n_pointer_only_licence":0}},{"rank_in_archive_order":3,"model":"EntNet","metrics":{"Accuracy (trained on 10k)":"99.5%","Accuracy (trained on 1k)":"89.1%","Mean Error Rate":"9.7%"},"uses_additional_data":false,"paper_date":"2016-12-12","paper":"/paper/tracking-the-world-state-with-recurrent","paper_url":"http://arxiv.org/abs/1612.03969v3","paper_title":"Tracking the World State with Recurrent Entity Networks","code":"https://github.com/facebook/MemNN","n_code_links":5,"syntology":{"n_ran":0,"n_unverified":3,"n_samples":3,"n_pointer_only_licence":0}},{"rank_in_archive_order":4,"model":"H-Mem","metrics":{"Accuracy (trained on 10k)":"99.4%","Accuracy (trained on 1k)":"77.1%"},"uses_additional_data":false,"paper_date":"2020-12-01","paper":"/paper/h-mem-harnessing-synaptic-plasticity-with","paper_url":"http://proceedings.neurips.cc/paper/2020/hash/f6876a9f998f6472cc26708e27444456-Abstract.html","paper_title":"H-Mem: Harnessing synaptic plasticity with Hebbian Memory Networks","code":"https://github.com/IGITUGraz/H-Mem","n_code_links":1,"syntology":null},{"rank_in_archive_order":5,"model":"DMN+","metrics":{"Accuracy (trained on 10k)":"97.2%","Accuracy (trained on 1k)":"66.8%"},"uses_additional_data":false,"paper_date":"2016-06-30","paper":"/paper/dynamic-neural-turing-machine-with-soft-and","paper_url":"http://arxiv.org/abs/1607.00036v2","paper_title":"Dynamic Neural Turing Machine with Soft and Hard Addressing Schemes","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":6,"model":"End-To-End Memory Networks","metrics":{"Accuracy (trained on 10k)":"93.4%","Accuracy (trained on 1k)":"86.1%","Mean Error Rate":" 7.5%"},"uses_additional_data":false,"paper_date":"2015-03-31","paper":"/paper/end-to-end-memory-networks","paper_url":"http://arxiv.org/abs/1503.08895v5","paper_title":"End-To-End Memory Networks","code":"https://github.com/PaddlePaddle/PaddleNLP/tree/develop/examples/language_model/end_to_end_memory_networks","n_code_links":44,"syntology":{"n_ran":2,"n_unverified":13,"n_samples":15,"n_pointer_only_licence":5}},{"rank_in_archive_order":7,"model":"ours","metrics":{"Accuracy (trained on 10k)":"89.1%"},"uses_additional_data":false,"paper_date":"2022-05-23","paper":"/paper/memory-enriched-computation-and-learning-in","paper_url":"https://arxiv.org/abs/2205.11276v1","paper_title":"Memory-enriched computation and learning in spiking neural networks through Hebbian plasticity","code":"https://github.com/igitugraz/memorydependentcomputation","n_code_links":1,"syntology":null},{"rank_in_archive_order":8,"model":"RUM","metrics":{"Accuracy (trained on 1k)":"73.2%"},"uses_additional_data":false,"paper_date":"2017-10-26","paper":"/paper/rotational-unit-of-memory","paper_url":"http://arxiv.org/abs/1710.09537v1","paper_title":"Rotational Unit of Memory","code":"https://github.com/jingli9111/RUM","n_code_links":2,"syntology":null},{"rank_in_archive_order":9,"model":"GORU","metrics":{"Accuracy (trained on 1k)":"60%"},"uses_additional_data":false,"paper_date":"2017-06-08","paper":"/paper/gated-orthogonal-recurrent-units-on-learning","paper_url":"http://arxiv.org/abs/1706.02761v3","paper_title":"Gated Orthogonal Recurrent Units: On Learning to Forget","code":"https://github.com/jingli9111/GORU-tensorflow","n_code_links":1,"syntology":null},{"rank_in_archive_order":10,"model":"LSTM","metrics":{"Accuracy (trained on 1k)":"49%","Mean Error Rate":"28.7%"},"uses_additional_data":false,"paper_date":"2016-10-27","paper":"/paper/scaling-memory-augmented-neural-networks-with","paper_url":"http://arxiv.org/abs/1610.09027v1","paper_title":"Scaling Memory-Augmented Neural Networks with Sparse Reads and Writes","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":11,"model":"RR","metrics":{"Mean Error Rate":"0.46%"},"uses_additional_data":false,"paper_date":"2017-11-21","paper":"/paper/recurrent-relational-networks","paper_url":"http://arxiv.org/abs/1711.08028v4","paper_title":"Recurrent Relational Networks","code":"https://github.com/dmlc/dgl/tree/master/examples/pytorch/rrn","n_code_links":6,"syntology":null},{"rank_in_archive_order":12,"model":"ReMO","metrics":{"Mean Error Rate":" 1.2%"},"uses_additional_data":false,"paper_date":"2018-01-25","paper":"/paper/finding-remo-related-memory-object-a-simple","paper_url":"http://arxiv.org/abs/1801.08459v2","paper_title":"Finding ReMO (Related Memory Object): A Simple Neural Architecture for Text based Reasoning","code":null,"n_code_links":0,"syntology":null},{"rank_in_archive_order":13,"model":"NUTM","metrics":{"Mean Error Rate":"5.6%"},"uses_additional_data":false,"paper_date":"2019-05-25","paper":"/paper/neural-stored-program-memory","paper_url":"https://arxiv.org/abs/1906.08862v2","paper_title":"Neural Stored-program Memory","code":"https://github.com/thaihungle/NSM","n_code_links":1,"syntology":{"n_ran":1,"n_unverified":3,"n_samples":4,"n_pointer_only_licence":0}},{"rank_in_archive_order":14,"model":"SDNC","metrics":{"Mean Error Rate":"6.4%"},"uses_additional_data":false,"paper_date":"2016-10-27","paper":"/paper/scaling-memory-augmented-neural-networks-with","paper_url":"http://arxiv.org/abs/1610.09027v1","paper_title":"Scaling Memory-Augmented Neural Networks with Sparse Reads and Writes","code":null,"n_code_links":0,"syntology":null}],"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":5,"rows_with_any_sample_ran":3,"distinct_papers_with_graph_line":5,"distinct_papers_with_any_sample_ran":3,"samples_over_distinct_papers":{"n_ran":4,"n_unverified":27,"n_samples":31,"n_pointer_only_licence":5,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":4,"n_unverified":27,"n_samples":31,"n_pointer_only_licence":5,"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"}}}