{"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/adaptive-posterior-learning-few-shot-learning","title":"Adaptive Posterior Learning: few-shot learning with a surprise-based memory module","arxiv_id":"1902.02527","date":"2019-02-07","proceeding":"ICLR 2019 5","authors":["Tiago Ramalho","Marta Garnelo"],"abstract":"The ability to generalize quickly from few observations is crucial for\nintelligent systems. In this paper we introduce APL, an algorithm that\napproximates probability distributions by remembering the most surprising\nobservations it has encountered. These past observations are recalled from an\nexternal memory module and processed by a decoder network that can combine\ninformation from different memory slots to generalize beyond direct recall. We\nshow this algorithm can perform as well as state of the art baselines on\nfew-shot classification benchmarks with a smaller memory footprint. In\naddition, its memory compression allows it to scale to thousands of unknown\nlabels. Finally, we introduce a meta-learning reasoning task which is more\nchallenging than direct classification. In this setting, APL is able to\ngeneralize with fewer than one example per class via deductive reasoning.","url_abs":"http://arxiv.org/abs/1902.02527v1","url_pdf":"http://arxiv.org/pdf/1902.02527v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"adaptive-posterior-learning-few-shot-learning","repo_url":"https://github.com/cogentlabs/apl","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"decoder","task_name":"Decoder"},{"task_slug":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"meta-learning","task_name":"Meta-Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/few-shot-image-classification-on-omniglot-1-5","task":"Few-Shot Image Classification","dataset":"OMNIGLOT - 1-Shot, 1000 way","model":"APL","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"68.9"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-omniglot-1-1","task":"Few-Shot Image Classification","dataset":"OMNIGLOT - 1-Shot, 20-way","model":"APL","rank_in_archive_order":8,"of":20,"metrics":{"Accuracy":"97.2%"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-omniglot-1-4","task":"Few-Shot Image Classification","dataset":"OMNIGLOT - 1-Shot, 423 way","model":"APL","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"73.5"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-omniglot-1-2","task":"Few-Shot Image Classification","dataset":"OMNIGLOT - 1-Shot, 5-way","model":"APL","rank_in_archive_order":16,"of":17,"metrics":{"Accuracy":"97.9"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-omniglot-5-5","task":"Few-Shot Image Classification","dataset":"OMNIGLOT - 5-Shot, 1000 way","model":"APL","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"78.9"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-omniglot-5-1","task":"Few-Shot Image Classification","dataset":"OMNIGLOT - 5-Shot, 20-way","model":"APL","rank_in_archive_order":17,"of":19,"metrics":{"Accuracy":"97.6%"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-omniglot-5-4","task":"Few-Shot Image Classification","dataset":"OMNIGLOT - 5-Shot, 423 way","model":"APL","rank_in_archive_order":1,"of":1,"metrics":{"Accuracy":"88"},"uses_additional_data":false},{"leaderboard":"/sota/few-shot-image-classification-on-omniglot-5-2","task":"Few-Shot Image Classification","dataset":"OMNIGLOT - 5-Shot, 5-way","model":"APL","rank_in_archive_order":3,"of":16,"metrics":{"Accuracy":"99.9"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1902.02527","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}