{"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":"/sota/sentiment-analysis-on-imdb","task":{"name":"Sentiment Analysis","url":"/task/sentiment-analysis","note":null},"dataset":{"name":"IMDb","url":"/dataset/imdb-movie-reviews"},"category":"Computer Vision","categories":["Computer Vision","Natural Language Processing"],"category_note":null,"description":"**Sentiment Analysis** is the task of classifying the polarity of a given text. For instance, a text-based tweet can be categorized into either \"positive\", \"negative\", or \"neutral\". Given the text and accompanying labels, a model can be trained to predict the correct sentiment. \r\n\r\n**Sentiment Analysis** techniques can be categorized into machine learning approaches, lexicon-based approaches, and even hybrid methods. Some subcategories of research in sentiment analysis include: multimodal sentiment analysis, aspect-based sentiment analysis, fine-grained opinion analysis, language specific sentiment analysis.\r\n\r\nMore recently, deep learning techniques, such as RoBERTa and T5, are used to train high-performing sentiment classifiers that are evaluated using metrics like F1, recall, and precision. To evaluate sentiment analysis systems, benchmark datasets like SST, GLUE, and IMDB movie reviews are used.\r\n\r\nFurther readings:\r\n\r\n- [Sentiment Analysis Based on Deep Learning: A Comparative Study](https://paperswithcode.com/paper/sentiment-analysis-based-on-deep-learning-a)","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"],"metric_direction":{"note":"inferred from the metric name only (the archive records no direction; the tokens 'score' and 'time' are not used on this page); null = not inferred, chart draws points only","by_metric":{"Accuracy":"higher"}},"counts":{"rows":49,"rows_with_code":47,"rows_with_paper_page":48,"rows_dated":48,"rows_using_additional_data":10},"rows":[{"rank_in_archive_order":1,"model":"RoBERTa-large with LlamBERT","metrics":{"Accuracy":"96.68"},"uses_additional_data":false,"paper_date":"2024-03-23","paper":"/paper/llambert-large-scale-low-cost-data-annotation","paper_url":"https://arxiv.org/abs/2403.15938v1","paper_title":"LlamBERT: Large-scale low-cost data annotation in NLP","code":"https://github.com/aielte-research/llambert","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":2,"model":"RoBERTa-large","metrics":{"Accuracy":"96.54"},"uses_additional_data":false,"paper_date":"2024-03-23","paper":"/paper/llambert-large-scale-low-cost-data-annotation","paper_url":"https://arxiv.org/abs/2403.15938v1","paper_title":"LlamBERT: Large-scale low-cost data annotation in NLP","code":"https://github.com/aielte-research/llambert","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":3,"model":"XLNet","metrics":{"Accuracy":"96.21"},"uses_additional_data":true,"paper_date":"2019-06-19","paper":"/paper/xlnet-generalized-autoregressive-pretraining","paper_url":"https://arxiv.org/abs/1906.08237v2","paper_title":"XLNet: Generalized Autoregressive Pretraining for Language Understanding","code":"https://github.com/huggingface/transformers","n_code_links":27,"official":{"repos":["zihangdai/xlnet"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":1,"ran_from_kinds":["listed","official"]},"syntology":{"n_ran":15,"n_constructed":2,"n_ran_checked":10,"n_instrument":5,"n_unverified":9,"n_samples":24,"n_pointer_only_licence":4,"sample_list":"/paper/xlnet-generalized-autoregressive-pretraining#ran","syntology_url":"https://syntology.ai/paper/1906.08237","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1906.08237"}}}},{"rank_in_archive_order":4,"model":"Heinsen Routing + RoBERTa Large","metrics":{"Accuracy":"96.2"},"uses_additional_data":false,"paper_date":"2022-11-20","paper":"/paper/an-algorithm-for-routing-vectors-in-sequences","paper_url":"https://arxiv.org/abs/2211.11754v3","paper_title":"An Algorithm for Routing Vectors in Sequences","code":"https://github.com/glassroom/heinsen_routing","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":5,"model":"RoBERTa-large 355M + Entailment as Few-shot Learner","metrics":{"Accuracy":"96.1"},"uses_additional_data":false,"paper_date":"2021-04-29","paper":"/paper/entailment-as-few-shot-learner","paper_url":"https://arxiv.org/abs/2104.14690v1","paper_title":"Entailment as Few-Shot Learner","code":"https://github.com/PaddlePaddle/PaddleNLP/tree/develop/examples/few_shot/efl","n_code_links":3,"official":null,"syntology":{"n_ran":1,"n_constructed":0,"n_ran_checked":1,"n_instrument":0,"n_unverified":2,"n_samples":3,"n_pointer_only_licence":0,"sample_list":"/paper/entailment-as-few-shot-learner#ran","syntology_url":"https://syntology.ai/paper/2104.14690","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2104.14690"}}}},{"rank_in_archive_order":6,"model":"GraphStar","metrics":{"Accuracy":"96.0"},"uses_additional_data":false,"paper_date":"2019-06-21","paper":"/paper/graph-star-net-for-generalized-multi-task-1","paper_url":"https://arxiv.org/abs/1906.12330v1","paper_title":"Graph Star Net for Generalized Multi-Task Learning","code":"https://github.com/graph-star-team/graph_star","n_code_links":1,"official":null,"syntology":{"n_ran":2,"n_constructed":0,"n_ran_checked":1,"n_instrument":1,"n_unverified":4,"n_samples":6,"n_pointer_only_licence":1,"sample_list":"/paper/graph-star-net-for-generalized-multi-task-1#ran","syntology_url":"https://syntology.ai/paper/1906.12330","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1906.12330"}}}},{"rank_in_archive_order":7,"model":"DV-ngrams-cosine with NB sub-sampling + RoBERTa.base","metrics":{"Accuracy":"95.94"},"uses_additional_data":false,"paper_date":"2022-05-26","paper":"/paper/the-document-vectors-using-cosine-similarity-1","paper_url":"https://arxiv.org/abs/2205.13357v1","paper_title":"The Document Vectors Using Cosine Similarity Revisited","code":"https://github.com/bgzh/dv_cosine_revisited","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":8,"model":"DV-ngrams-cosine + RoBERTa.base","metrics":{"Accuracy":"95.92"},"uses_additional_data":false,"paper_date":"2022-05-26","paper":"/paper/the-document-vectors-using-cosine-similarity-1","paper_url":"https://arxiv.org/abs/2205.13357v1","paper_title":"The Document Vectors Using Cosine Similarity Revisited","code":"https://github.com/bgzh/dv_cosine_revisited","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":9,"model":"Roberta_Large ST + Cosine Similarity Loss","metrics":{"Accuracy":"95.9"},"uses_additional_data":false,"paper_date":null,"paper":null,"paper_url":null,"paper_title":"","code":null,"n_code_links":0,"official":null,"syntology":null},{"rank_in_archive_order":10,"model":"BERT large finetune UDA","metrics":{"Accuracy":"95.8"},"uses_additional_data":true,"paper_date":"2019-04-29","paper":"/paper/unsupervised-data-augmentation-1","paper_url":"https://arxiv.org/abs/1904.12848v6","paper_title":"Unsupervised Data Augmentation for Consistency Training","code":"https://github.com/google-research/uda","n_code_links":20,"official":{"repos":["google-research/uda"],"state":"official (archive's flag): 10 ran","n_ran":10,"n_constructed":0,"n_ran_no_instrument_failure":10,"n_unverified":14,"ran_from_kinds":["listed","official"]},"syntology":{"n_ran":30,"n_constructed":3,"n_ran_checked":22,"n_instrument":8,"n_unverified":22,"n_samples":52,"n_pointer_only_licence":17,"sample_list":"/paper/unsupervised-data-augmentation-1#ran","syntology_url":"https://syntology.ai/paper/1904.12848","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1904.12848"}}}},{"rank_in_archive_order":11,"model":"BERT_large+ITPT","metrics":{"Accuracy":"95.79"},"uses_additional_data":false,"paper_date":"2019-05-14","paper":"/paper/how-to-fine-tune-bert-for-text-classification","paper_url":"https://arxiv.org/abs/1905.05583v3","paper_title":"How to Fine-Tune BERT for Text Classification?","code":"https://github.com/xuyige/BERT4doc-Classification","n_code_links":15,"official":{"repos":["xuyige/BERT4doc-Classification"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["listed","official"]},"syntology":{"n_ran":12,"n_constructed":0,"n_ran_checked":7,"n_instrument":5,"n_unverified":6,"n_samples":18,"n_pointer_only_licence":5,"sample_list":"/paper/how-to-fine-tune-bert-for-text-classification#ran","syntology_url":"https://syntology.ai/paper/1905.05583","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1905.05583"}}}},{"rank_in_archive_order":12,"model":"RoBERTa.base","metrics":{"Accuracy":"95.79"},"uses_additional_data":true,"paper_date":"2022-05-26","paper":"/paper/the-document-vectors-using-cosine-similarity-1","paper_url":"https://arxiv.org/abs/2205.13357v1","paper_title":"The Document Vectors Using Cosine Similarity Revisited","code":"https://github.com/bgzh/dv_cosine_revisited","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":13,"model":"L MIXED","metrics":{"Accuracy":"95.68"},"uses_additional_data":false,"paper_date":"2020-09-08","paper":"/paper/revisiting-lstm-networks-for-semi-supervised-1","paper_url":"https://arxiv.org/abs/2009.04007v1","paper_title":"Revisiting LSTM Networks for Semi-Supervised Text Classification via Mixed Objective Function","code":"https://github.com/DevSinghSachan/ssl_text_classification","n_code_links":1,"official":null,"syntology":{"n_ran":2,"n_constructed":0,"n_ran_checked":0,"n_instrument":2,"n_unverified":1,"n_samples":3,"n_pointer_only_licence":3,"sample_list":"/paper/revisiting-lstm-networks-for-semi-supervised-1#ran","syntology_url":"https://syntology.ai/paper/2009.04007","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2009.04007"}}}},{"rank_in_archive_order":14,"model":"BERT_base+ITPT","metrics":{"Accuracy":"95.63"},"uses_additional_data":false,"paper_date":"2019-05-14","paper":"/paper/how-to-fine-tune-bert-for-text-classification","paper_url":"https://arxiv.org/abs/1905.05583v3","paper_title":"How to Fine-Tune BERT for Text Classification?","code":"https://github.com/xuyige/BERT4doc-Classification","n_code_links":15,"official":{"repos":["xuyige/BERT4doc-Classification"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["listed","official"]},"syntology":{"n_ran":12,"n_constructed":0,"n_ran_checked":7,"n_instrument":5,"n_unverified":6,"n_samples":18,"n_pointer_only_licence":5,"sample_list":"/paper/how-to-fine-tune-bert-for-text-classification#ran","syntology_url":"https://syntology.ai/paper/1905.05583","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1905.05583"}}}},{"rank_in_archive_order":15,"model":"BERT large","metrics":{"Accuracy":"95.49"},"uses_additional_data":false,"paper_date":"2019-04-29","paper":"/paper/unsupervised-data-augmentation-1","paper_url":"https://arxiv.org/abs/1904.12848v6","paper_title":"Unsupervised Data Augmentation for Consistency Training","code":"https://github.com/google-research/uda","n_code_links":20,"official":{"repos":["google-research/uda"],"state":"official (archive's flag): 10 ran","n_ran":10,"n_constructed":0,"n_ran_no_instrument_failure":10,"n_unverified":14,"ran_from_kinds":["listed","official"]},"syntology":{"n_ran":30,"n_constructed":3,"n_ran_checked":22,"n_instrument":8,"n_unverified":22,"n_samples":52,"n_pointer_only_licence":17,"sample_list":"/paper/unsupervised-data-augmentation-1#ran","syntology_url":"https://syntology.ai/paper/1904.12848","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1904.12848"}}}},{"rank_in_archive_order":16,"model":"ULMFiT","metrics":{"Accuracy":"95.4"},"uses_additional_data":false,"paper_date":"2018-01-18","paper":"/paper/universal-language-model-fine-tuning-for-text","paper_url":"http://arxiv.org/abs/1801.06146v5","paper_title":"Universal Language Model Fine-tuning for Text Classification","code":"https://github.com/fastai/fastai","n_code_links":66,"official":{"repos":["fastai/fastai"],"state":"community repositories only","n_ran":0,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":["listed"]},"syntology":{"n_ran":2,"n_constructed":0,"n_ran_checked":0,"n_instrument":2,"n_unverified":3,"n_samples":5,"n_pointer_only_licence":3,"sample_list":"/paper/universal-language-model-fine-tuning-for-text#ran","syntology_url":"https://syntology.ai/paper/1801.06146","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1801.06146"}}}},{"rank_in_archive_order":17,"model":"Llama-2-70b-chat (0-shot)","metrics":{"Accuracy":"95.39"},"uses_additional_data":true,"paper_date":"2024-03-23","paper":"/paper/llambert-large-scale-low-cost-data-annotation","paper_url":"https://arxiv.org/abs/2403.15938v1","paper_title":"LlamBERT: Large-scale low-cost data annotation in NLP","code":"https://github.com/aielte-research/llambert","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":18,"model":"FLAN 137B (few-shot, k=2)","metrics":{"Accuracy":"95"},"uses_additional_data":true,"paper_date":"2021-09-03","paper":"/paper/finetuned-language-models-are-zero-shot","paper_url":"https://arxiv.org/abs/2109.01652v5","paper_title":"Finetuned Language Models Are Zero-Shot Learners","code":"https://github.com/hiyouga/llama-efficient-tuning","n_code_links":8,"official":{"repos":["google-research/flan"],"state":"official: harvested for another paper","n_ran":0,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":[]},"syntology":{"n_ran":0,"n_constructed":0,"n_ran_checked":0,"n_instrument":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0,"sample_list":"/paper/finetuned-language-models-are-zero-shot#ran","syntology_url":"https://syntology.ai/paper/2109.01652","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2109.01652"}}}},{"rank_in_archive_order":19,"model":"Block-sparse LSTM","metrics":{"Accuracy":"94.99"},"uses_additional_data":false,"paper_date":"2017-12-01","paper":"/paper/gpu-kernels-for-block-sparse-weights","paper_url":"https://blog.openai.com/block-sparse-gpu-kernels/","paper_title":"GPU Kernels for Block-Sparse Weights","code":"https://github.com/openai/blocksparse","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":20,"model":"Space-XLNet","metrics":{"Accuracy":"94.88"},"uses_additional_data":false,"paper_date":"2024-01-30","paper":"/paper/breaking-free-transformer-models-task","paper_url":"https://arxiv.org/abs/2401.16638v1","paper_title":"Breaking Free Transformer Models: Task-specific Context Attribution Promises Improved Generalizability Without Fine-tuning Pre-trained LLMs","code":"https://github.com/stepantita/space-model","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":21,"model":"CEN-tpc","metrics":{"Accuracy":"94.52"},"uses_additional_data":false,"paper_date":"2017-05-29","paper":"/paper/contextual-explanation-networks","paper_url":"https://arxiv.org/abs/1705.10301v4","paper_title":"Contextual Explanation Networks","code":"https://github.com/alshedivat/cen","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":22,"model":"FLAN 137B (zero-shot)","metrics":{"Accuracy":"94.3"},"uses_additional_data":true,"paper_date":"2021-09-03","paper":"/paper/finetuned-language-models-are-zero-shot","paper_url":"https://arxiv.org/abs/2109.01652v5","paper_title":"Finetuned Language Models Are Zero-Shot Learners","code":"https://github.com/hiyouga/llama-efficient-tuning","n_code_links":8,"official":{"repos":["google-research/flan"],"state":"official: harvested for another paper","n_ran":0,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":[]},"syntology":{"n_ran":0,"n_constructed":0,"n_ran_checked":0,"n_instrument":0,"n_unverified":1,"n_samples":1,"n_pointer_only_licence":0,"sample_list":"/paper/finetuned-language-models-are-zero-shot#ran","syntology_url":"https://syntology.ai/paper/2109.01652","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2109.01652"}}}},{"rank_in_archive_order":23,"model":"oh-LSTM","metrics":{"Accuracy":"94.1"},"uses_additional_data":true,"paper_date":"2016-02-07","paper":"/paper/supervised-and-semi-supervised-text","paper_url":"http://arxiv.org/abs/1602.02373v2","paper_title":"Supervised and Semi-Supervised Text Categorization using LSTM for Region Embeddings","code":null,"n_code_links":0,"official":null,"syntology":null},{"rank_in_archive_order":24,"model":"Virtual adversarial training","metrics":{"Accuracy":"94.1"},"uses_additional_data":false,"paper_date":"2016-05-25","paper":"/paper/adversarial-training-methods-for-semi","paper_url":"https://arxiv.org/abs/1605.07725v4","paper_title":"Adversarial Training Methods for Semi-Supervised Text Classification","code":"https://github.com/tensorflow/models/tree/master/research/adversarial_text","n_code_links":4,"official":null,"syntology":null},{"rank_in_archive_order":25,"model":"DV-ngrams-cosine + NB-weighted BON (re-evaluated)","metrics":{"Accuracy":"93.68"},"uses_additional_data":false,"paper_date":"2022-05-26","paper":"/paper/the-document-vectors-using-cosine-similarity-1","paper_url":"https://arxiv.org/abs/2205.13357v1","paper_title":"The Document Vectors Using Cosine Similarity Revisited","code":"https://github.com/bgzh/dv_cosine_revisited","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":26,"model":"Nyströmformer","metrics":{"Accuracy":"93.2"},"uses_additional_data":false,"paper_date":"2021-02-07","paper":"/paper/nystromformer-a-nystrom-based-algorithm-for","paper_url":"https://arxiv.org/abs/2102.03902v3","paper_title":"Nyströmformer: A Nyström-Based Algorithm for Approximating Self-Attention","code":"https://github.com/facebookresearch/xformers","n_code_links":10,"official":{"repos":["mlpen/Nystromformer"],"state":"official: no sample here; runs from other or unrecorded repositories","n_ran":0,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":["unlocated"]},"syntology":{"n_ran":1,"n_constructed":0,"n_ran_checked":1,"n_instrument":0,"n_unverified":1,"n_samples":2,"n_pointer_only_licence":1,"sample_list":"/paper/nystromformer-a-nystrom-based-algorithm-for#ran","syntology_url":"https://syntology.ai/paper/2102.03902","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2102.03902"}}}},{"rank_in_archive_order":27,"model":"Modified LMU (34M)","metrics":{"Accuracy":"93.20"},"uses_additional_data":false,"paper_date":"2021-02-22","paper":"/paper/parallelizing-legendre-memory-unit-training","paper_url":"https://arxiv.org/abs/2102.11417v2","paper_title":"Parallelizing Legendre Memory Unit Training","code":"https://github.com/hrshtv/pytorch-lmu","n_code_links":2,"official":null,"syntology":null},{"rank_in_archive_order":28,"model":"DV-ngrams-cosine","metrics":{"Accuracy":"93.13"},"uses_additional_data":false,"paper_date":"2019-07-01","paper":"/paper/sentiment-classification-using-document","paper_url":"https://aclanthology.org/P19-2057","paper_title":"Sentiment Classification Using Document Embeddings Trained with Cosine Similarity","code":"https://github.com/tanthongtan/dv-cosine","n_code_links":2,"official":null,"syntology":null},{"rank_in_archive_order":29,"model":"OCaTS (kNN & GPT-3.5-turbo","metrics":{"Accuracy":"93.06"},"uses_additional_data":false,"paper_date":"2023-10-20","paper":"/paper/cache-me-if-you-can-an-online-cost-aware","paper_url":"https://arxiv.org/abs/2310.13395v1","paper_title":"Cache me if you Can: an Online Cost-aware Teacher-Student framework to Reduce the Calls to Large Language Models","code":"https://github.com/stoyian/OCaTS","n_code_links":1,"official":{"repos":["stoyian/OCaTS"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["official"]},"syntology":{"n_ran":1,"n_constructed":0,"n_ran_checked":1,"n_instrument":0,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":0,"sample_list":"/paper/cache-me-if-you-can-an-online-cost-aware#ran","syntology_url":"https://syntology.ai/paper/2310.13395","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2310.13395"}}}},{"rank_in_archive_order":30,"model":"DistilBERT 66M","metrics":{"Accuracy":"92.82"},"uses_additional_data":false,"paper_date":"2019-10-02","paper":"/paper/distilbert-a-distilled-version-of-bert","paper_url":"https://arxiv.org/abs/1910.01108v4","paper_title":"DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter","code":"https://github.com/huggingface/transformers","n_code_links":37,"official":{"repos":["huggingface/swift-coreml-transformers","huggingface/transformers"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["listed","official"]},"syntology":{"n_ran":21,"n_constructed":5,"n_ran_checked":12,"n_instrument":9,"n_unverified":6,"n_samples":27,"n_pointer_only_licence":2,"sample_list":"/paper/distilbert-a-distilled-version-of-bert#ran","syntology_url":"https://syntology.ai/paper/1910.01108","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1910.01108"}}}},{"rank_in_archive_order":31,"model":"GPT-2 Finetuned","metrics":{"Accuracy":"92.36"},"uses_additional_data":false,"paper_date":"2019-02-14","paper":"/paper/language-models-are-unsupervised-multitask","paper_url":"https://d4mucfpksywv.cloudfront.net/better-language-models/language-models.pdf","paper_title":"Language Models are Unsupervised Multitask Learners","code":"https://github.com/huggingface/transformers","n_code_links":21,"official":null,"syntology":null},{"rank_in_archive_order":32,"model":"seq2-bown-CNN","metrics":{"Accuracy":"92.33"},"uses_additional_data":false,"paper_date":"2014-12-01","paper":"/paper/effective-use-of-word-order-for-text-1","paper_url":"http://arxiv.org/abs/1412.1058v2","paper_title":"Effective Use of Word Order for Text Categorization with Convolutional Neural Networks","code":"https://github.com/tensorflow/models","n_code_links":4,"official":null,"syntology":null},{"rank_in_archive_order":33,"model":"BP-Transformer + GloVe","metrics":{"Accuracy":"92.12"},"uses_additional_data":false,"paper_date":"2019-11-11","paper":"/paper/bp-transformer-modelling-long-range-context","paper_url":"https://arxiv.org/abs/1911.04070v1","paper_title":"BP-Transformer: Modelling Long-Range Context via Binary Partitioning","code":"https://github.com/dmlc/dgl/tree/master/examples/pytorch/transformer","n_code_links":2,"official":{"repos":["yzh119/BPT"],"state":"official (archive's flag): 2 ran","n_ran":2,"n_constructed":0,"n_ran_no_instrument_failure":2,"n_unverified":5,"ran_from_kinds":["official"]},"syntology":{"n_ran":2,"n_constructed":0,"n_ran_checked":2,"n_instrument":0,"n_unverified":5,"n_samples":7,"n_pointer_only_licence":0,"sample_list":"/paper/bp-transformer-modelling-long-range-context#ran","syntology_url":"https://syntology.ai/paper/1911.04070","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1911.04070"}}}},{"rank_in_archive_order":34,"model":"BCN+Char+CoVe","metrics":{"Accuracy":"91.8"},"uses_additional_data":false,"paper_date":"2017-08-01","paper":"/paper/learned-in-translation-contextualized-word","paper_url":"http://arxiv.org/abs/1708.00107v2","paper_title":"Learned in Translation: Contextualized Word Vectors","code":"https://github.com/salesforce/cove","n_code_links":5,"official":{"repos":["salesforce/cove"],"state":"community repositories only","n_ran":0,"n_constructed":0,"n_ran_no_instrument_failure":0,"n_unverified":0,"ran_from_kinds":["listed"]},"syntology":{"n_ran":3,"n_constructed":0,"n_ran_checked":2,"n_instrument":1,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":2,"sample_list":"/paper/learned-in-translation-contextualized-word#ran","syntology_url":"https://syntology.ai/paper/1708.00107","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1708.00107"}}}},{"rank_in_archive_order":35,"model":"ToWE-SG","metrics":{"Accuracy":"90.8"},"uses_additional_data":false,"paper_date":"2018-08-01","paper":"/paper/task-oriented-word-embedding-for-text","paper_url":"https://aclanthology.org/C18-1172","paper_title":"Task-oriented Word Embedding for Text Classification","code":"https://github.com/qianliu0708/ToWE","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":36,"model":"COSINE","metrics":{"Accuracy":"90.54"},"uses_additional_data":false,"paper_date":"2020-10-15","paper":"/paper/fine-tuning-pre-trained-language-model-with","paper_url":"https://arxiv.org/abs/2010.07835v3","paper_title":"Fine-Tuning Pre-trained Language Model with Weak Supervision: A Contrastive-Regularized Self-Training Approach","code":"https://github.com/yueyu1030/COSINE","n_code_links":1,"official":{"repos":["yueyu1030/COSINE"],"state":"official (archive's flag): 5 ran","n_ran":5,"n_constructed":0,"n_ran_no_instrument_failure":5,"n_unverified":0,"ran_from_kinds":["official"]},"syntology":{"n_ran":5,"n_constructed":0,"n_ran_checked":5,"n_instrument":0,"n_unverified":0,"n_samples":5,"n_pointer_only_licence":0,"sample_list":"/paper/fine-tuning-pre-trained-language-model-with#ran","syntology_url":"https://syntology.ai/paper/2010.07835","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2010.07835"}}}},{"rank_in_archive_order":37,"model":"LSTM with dynamic skip","metrics":{"Accuracy":"90.1"},"uses_additional_data":false,"paper_date":"2018-11-09","paper":"/paper/long-short-term-memory-with-dynamic-skip","paper_url":"http://arxiv.org/abs/1811.03873v1","paper_title":"Long Short-Term Memory with Dynamic Skip Connections","code":"https://github.com/lecholin/DynamicLSTM","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":38,"model":"CNN+LSTM","metrics":{"Accuracy":"88.9"},"uses_additional_data":true,"paper_date":"2017-07-06","paper":"/paper/on-the-role-of-text-preprocessing-in-neural","paper_url":"http://arxiv.org/abs/1707.01780v3","paper_title":"On the Role of Text Preprocessing in Neural Network Architectures: An Evaluation Study on Text Categorization and Sentiment Analysis","code":"https://github.com/pedrada88/preproc-textclassification","n_code_links":3,"official":null,"syntology":null},{"rank_in_archive_order":39,"model":"UnICORNN","metrics":{"Accuracy":"88.4"},"uses_additional_data":false,"paper_date":"2021-03-09","paper":"/paper/unicornn-a-recurrent-model-for-learning-very","paper_url":"https://arxiv.org/abs/2103.05487v2","paper_title":"UnICORNN: A recurrent model for learning very long time dependencies","code":"https://github.com/tk-rusch/unicornn","n_code_links":1,"official":{"repos":["tk-rusch/unicornn"],"state":"official (archive's flag): 1 ran","n_ran":1,"n_constructed":0,"n_ran_no_instrument_failure":1,"n_unverified":0,"ran_from_kinds":["official"]},"syntology":{"n_ran":1,"n_constructed":0,"n_ran_checked":1,"n_instrument":0,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1,"sample_list":"/paper/unicornn-a-recurrent-model-for-learning-very#ran","syntology_url":"https://syntology.ai/paper/2103.05487","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2103.05487"}}}},{"rank_in_archive_order":40,"model":"CfC","metrics":{"Accuracy":"88.4"},"uses_additional_data":false,"paper_date":"2021-06-25","paper":"/paper/closed-form-continuous-depth-models","paper_url":"https://arxiv.org/abs/2106.13898v2","paper_title":"Closed-form Continuous-time Neural Models","code":"https://github.com/raminmh/CfC","n_code_links":1,"official":{"repos":["raminmh/CfC"],"state":"official (archive's flag): 7 ran","n_ran":7,"n_constructed":0,"n_ran_no_instrument_failure":5,"n_unverified":3,"ran_from_kinds":["official"]},"syntology":{"n_ran":7,"n_constructed":0,"n_ran_checked":5,"n_instrument":2,"n_unverified":3,"n_samples":10,"n_pointer_only_licence":1,"sample_list":"/paper/closed-form-continuous-depth-models#ran","syntology_url":"https://syntology.ai/paper/2106.13898","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2106.13898"}}}},{"rank_in_archive_order":41,"model":"Doc2VecC","metrics":{"Accuracy":"88.3"},"uses_additional_data":false,"paper_date":"2017-07-08","paper":"/paper/efficient-vector-representation-for-documents","paper_url":"http://arxiv.org/abs/1707.02377v1","paper_title":"Efficient Vector Representation for Documents through Corruption","code":"https://github.com/mchen24/iclr2017","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":42,"model":"Bert+ Wilson-Cowan model RNN","metrics":{"Accuracy":"87.46"},"uses_additional_data":false,"paper_date":"2024-06-24","paper":"/paper/learning-in-wilson-cowan-model-for","paper_url":"https://arxiv.org/abs/2406.16453v2","paper_title":"Learning in Wilson-Cowan model for metapopulation","code":"https://github.com/raffaelemarino/learning_in_wilsoncowan","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":43,"model":"coRNN","metrics":{"Accuracy":"87.4%"},"uses_additional_data":false,"paper_date":"2020-10-02","paper":"/paper/coupled-oscillatory-recurrent-neural-network","paper_url":"https://arxiv.org/abs/2010.00951v2","paper_title":"Coupled Oscillatory Recurrent Neural Network (coRNN): An accurate and (gradient) stable architecture for learning long time dependencies","code":"https://github.com/tk-rusch/coRNN","n_code_links":1,"official":{"repos":["tk-rusch/coRNN"],"state":"official (archive's flag): 3 ran","n_ran":3,"n_constructed":0,"n_ran_no_instrument_failure":3,"n_unverified":0,"ran_from_kinds":["official"]},"syntology":{"n_ran":3,"n_constructed":0,"n_ran_checked":3,"n_instrument":0,"n_unverified":0,"n_samples":3,"n_pointer_only_licence":1,"sample_list":"/paper/coupled-oscillatory-recurrent-neural-network#ran","syntology_url":"https://syntology.ai/paper/2010.00951","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2010.00951"}}}},{"rank_in_archive_order":44,"model":"S-LSTM","metrics":{"Accuracy":"87.15"},"uses_additional_data":true,"paper_date":"2018-05-07","paper":"/paper/sentence-state-lstm-for-text-representation","paper_url":"http://arxiv.org/abs/1805.02474v1","paper_title":"Sentence-State LSTM for Text Representation","code":"https://github.com/leuchine/S-LSTM","n_code_links":2,"official":null,"syntology":{"n_ran":1,"n_constructed":0,"n_ran_checked":0,"n_instrument":1,"n_unverified":0,"n_samples":1,"n_pointer_only_licence":1,"sample_list":"/paper/sentence-state-lstm-for-text-representation#ran","syntology_url":"https://syntology.ai/paper/1805.02474","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1805.02474"}}}},{"rank_in_archive_order":45,"model":"AlexNet [alexnet]","metrics":{"Accuracy":"87"},"uses_additional_data":false,"paper_date":"2021-06-23","paper":"/paper/classifying-textual-data-with-pre-trained","paper_url":"https://arxiv.org/abs/2106.12479v4","paper_title":"Classifying Textual Data with Pre-trained Vision Models through Transfer Learning and Data Transformations","code":"https://github.com/EddCBen/Classifying-Textual-Data-with-pretrained-Vision-Models-through-Transfer-Learning-and-Data-Transforms","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":46,"model":"VGG16 [vgg16]","metrics":{"Accuracy":"86"},"uses_additional_data":false,"paper_date":"2021-06-23","paper":"/paper/classifying-textual-data-with-pre-trained","paper_url":"https://arxiv.org/abs/2106.12479v4","paper_title":"Classifying Textual Data with Pre-trained Vision Models through Transfer Learning and Data Transformations","code":"https://github.com/EddCBen/Classifying-Textual-Data-with-pretrained-Vision-Models-through-Transfer-Learning-and-Data-Transforms","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":47,"model":"ResNext[resnext]","metrics":{"Accuracy":"85"},"uses_additional_data":false,"paper_date":"2021-06-23","paper":"/paper/classifying-textual-data-with-pre-trained","paper_url":"https://arxiv.org/abs/2106.12479v4","paper_title":"Classifying Textual Data with Pre-trained Vision Models through Transfer Learning and Data Transformations","code":"https://github.com/EddCBen/Classifying-Textual-Data-with-pretrained-Vision-Models-through-Transfer-Learning-and-Data-Transforms","n_code_links":1,"official":null,"syntology":null},{"rank_in_archive_order":48,"model":"Standard DR-AGG","metrics":{"Accuracy":"45.1"},"uses_additional_data":true,"paper_date":"2018-06-05","paper":"/paper/information-aggregation-via-dynamic-routing","paper_url":"http://arxiv.org/abs/1806.01501v1","paper_title":"Information Aggregation via Dynamic Routing for Sequence Encoding","code":"https://github.com/FudanNLP/Capsule4TextClassification","n_code_links":2,"official":null,"syntology":null},{"rank_in_archive_order":49,"model":"Reverse DR-AGG","metrics":{"Accuracy":"44.5"},"uses_additional_data":false,"paper_date":"2018-06-05","paper":"/paper/information-aggregation-via-dynamic-routing","paper_url":"http://arxiv.org/abs/1806.01501v1","paper_title":"Information Aggregation via Dynamic Routing for Sequence Encoding","code":"https://github.com/FudanNLP/Capsule4TextClassification","n_code_links":2,"official":null,"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,"adjudicator":"Claude Opus 5.5","adjudicator_from":"groundwork/59_code_with_papers/LB_EXTRACTOR_PREREG.md (not recorded in lb_measurement.json; the file's frozen_commit, measured_on and counts equal the run the pre-registration records)","adjudicator_sensitivity":[69,70],"adjudicator_specificity":[101,103],"measured_on":"2026-09-24","frozen_commit":"0e3de0df94"},"measurement_note":"blind adjudication by a model (adjudicator) of accepted entries on a held-out split of archive papers, rules frozen before the test; adjudicator_sensitivity and adjudicator_specificity are its own controls as [judged right, of]; it was not measured on the newer papers placed here","measurement_sentence":"Measured before publishing, on held-out archive papers: the extractor was given 883 of them and produced output for 881 (the other 2 had no title and were excluded before any model call, so they are not part of this measurement). A model adjudicator (Claude Opus 5.5), blind to which entries were real, judged 108 of 110 accepted entries correct on those held-out archive papers; its own controls: 69 of 70 archive-confirmed entries judged true, 101 of 103 corrupted copies judged false. It was not measured on the newer papers whose results are placed on this site.","coverage":{"sentence":"Syntology has checked 8,886 of the 18,637 papers on this site that are newer than the archive (for 3,717 of them no archive leaderboard matched the paper's tables, so there was nothing further to check); 733 were read and have nothing to place (arXiv has no HTML version of the paper, or that version has no tables), 41 could not be read (the extractor's reply could not be parsed), and results from the other 8,977 appear after they are checked.","complete":false,"papers_newer_than_archive":18637,"papers_checked":8886,"papers_checked_no_leaderboard_matched":3717,"papers_read_nothing_to_place":733,"papers_could_not_be_read":41,"papers_could_not_be_read_by_status":{"unparseable":41},"papers_extracted_not_yet_verified":0,"boards_without_verdict":3,"papers_not_yet_extracted":8975},"board_status":{"sentence":"Of the 8,888 newer papers whose tables Syntology has read (8,886 checked in full and 2 with a leaderboard still awaiting a verdict), 25 had this board among the leaderboards the extractor considered for their tables (it keeps at most 10 boards per paper and 4 per dataset; the lists of 1,579 of the 8,888 were already full without this board, so if it was crowded out of them it is not counted), and 3 proposed a result; of those 3, 1 gave no cell that passed the extractor's own checks, 2 were rejected by the independent check, 0 await a verdict and 0 were placed.","note":"papers, from the extractor's own records, over the newer papers whose tables were read; proposed = no_cell_passed_checks + rejected_by_independent_check + accepted_not_placed + awaiting_verdict + placed; counts, never a rate","eligible":true,"eligible_note":"false when the extractor never considers this board: it loads only boards with archive rows and at least one metric column (lb_extract.py Boards), so no newer paper can be proposed here; archive_rows and metric_columns say which","archive_rows":49,"metric_columns":1,"papers_read":8888,"papers_read_checked":8886,"papers_read_awaiting_verdict":2,"papers_read_verified":8888,"papers_read_not_yet_verified":0,"papers_read_at_candidate_cap":1596,"papers_read_list_full_without_this_board":1579,"candidates":25,"candidates_capped_note":"candidates = papers that had this board among the leaderboards the extractor considered for their tables; the extractor keeps at most 10 per paper and 4 per dataset, so a board crowded out of a paper's list is not counted and candidates is a floor; papers_read_list_full_without_this_board counts the read papers whose list did not hold this board and had reached 10, or held 4 boards of this board's dataset (the only papers that could have crowded it out); papers_read_at_candidate_cap counts, over every board, the papers whose list reached 10","proposed":3,"no_cell_passed_checks":1,"rejected_by_independent_check":2,"accepted_not_placed":0,"awaiting_verdict":0,"placed":0,"placed_entries_listed":0,"file_lines_unreadable_as_read":0,"since_line":"Since the archive (Syntology): none placed; 3 newer papers judged to report on this board so far"},"judged_to_report":{"heading":"Newer papers whose tables Claude Sonnet 4.5 judged to report on this benchmark (a model's judgement, not a result; newest first by arXiv id)","judged_by":{"model":"Claude Sonnet 4.5","models_as_recorded":["global.anthropic.claude-sonnet-4-5-20250929-v1:0"],"a_record_without_a_model":false,"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","order":"newest first by arXiv id (the identifier's order, not a rank)","outcomes":{"placed":"placed","no_cell_passed_checks":"no cell passed the extractor's checks","refused_by_rule":"refused by a rule","rejected":"rejected by the independent check","accepted_not_placed":"accepted by the independent check but not placed by this site","awaiting":"awaiting a verdict"},"reason_note":"for refused_by_rule, the rules' plain words for the reason codes the record gives (never the quoted or the actual cell text); null for every other outcome","total":3,"shown_on_page":3,"no_cell_split":{"no_cell_passed_checks_in_ledger":1,"rows_refused_by_rule":1,"rows_no_cell_passed_checks":0},"rows":[{"arxiv_id":"2606.10610","outcome":"refused_by_rule","reason":"the model pointed at a header row; the model gave no value to check against the cell"},{"arxiv_id":"2603.05690","outcome":"rejected","reason":null},{"arxiv_id":"2602.02708","outcome":"rejected","reason":null}]},"order":"newest first by month (arXiv date, else the arXiv-id month), then arXiv id descending","columns":[],"entries":[]},"syntology":{"read_at":"2026-10-05T00:36:11+00:00","claim":"Per row: of M harvested code samples from that row's paper, n_ran executed on a synthesized fixture (n_ran = n_ran_checked + n_instrument; n_instrument counts failures of Syntology's instrument, not of the code); the other M - n_ran 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).","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)"},"rows_with_graph_line":21,"rows_with_any_sample_ran":19,"rows_with_a_run_with_no_instrument_failure":16,"official":{"state_rule":"the archive's official repositories have samples for this paper that ran: 'official (archive's flag): N ran' (N counts them, instrument failures included); they have samples, none ran: 'official: harvested, nothing ran'; they have no sample for this paper and another sample ran: 'community repositories only' when every sample that ran came from a repository the archive lists for the paper but neither marks official nor records as mentioned in it, or one Syntology harvested outside the archive's links, else 'official: no sample here; runs from other or unrecorded repositories' (a repository named in the paper or found in its text by Syntology, or no repository recorded); otherwise 'official: harvested for another paper' or 'official: not harvested'. Absent when the archive marks no repository official. Official is the archive's is_official flag, not an authorship verdict","flag":"the archive's is_official flag for the paper's code link; not a verdict on who wrote the code","rows_with_an_archive_official_repository":17,"rows_whose_official_repository_has_a_sample_that_ran":12,"rows_whose_official_repository_has_a_run_with_no_instrument_failure":12,"rows_community_repositories_only":2,"rows_runs_from_other_or_unrecorded_repositories":1,"note":"rows[].official, per row with a graph line: repos (marked official in the archive), state, n_ran and n_unverified (samples from those repositories), and ran_from_kinds (the kinds of repository the paper's samples that ran came from; 'unlocated' is no repository recorded)"},"filters":{"ran_modes":["a run with no instrument failure","any run, instrument failures included"],"default":"a run with no instrument failure","note":"the page's filters hide rows and never renumber them; they read the Ran cell's counts, the same figures rows[].syntology and rows[].official carry"},"row_route":"rows[].syntology 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","distinct_papers_with_graph_line":18,"distinct_papers_with_any_sample_ran":17,"distinct_papers_with_a_run_with_no_instrument_failure":14,"samples_over_distinct_papers":{"n_ran":109,"n_constructed":10,"n_ran_checked":73,"n_instrument":36,"n_unverified":63,"n_samples":172,"n_pointer_only_licence":42,"note":"each paper (arXiv id) counted once, however many rows it is behind; this is the page-level figure"},"samples_row_weighted":{"n_ran":151,"n_constructed":13,"n_ran_checked":102,"n_instrument":49,"n_unverified":92,"n_samples":243,"n_pointer_only_licence":64,"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"}},"record_sha256":"d77f014f2a947ea07878bb6c739d9c338695e696c32140317a8bc42cc52fbcb1","record_changed_at":"2026-10-05","record_changed_at_basis":"changed"}