{"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/meta-learning-with-warped-gradient-descent","title":"Meta-Learning with Warped Gradient Descent","arxiv_id":"1909.00025","date":"2019-08-30","proceeding":"ICLR 2020 1","authors":["Sebastian Flennerhag","Andrei A. Rusu","Razvan Pascanu","Francesco Visin","Hujun Yin","Raia Hadsell"],"abstract":"Learning an efficient update rule from data that promotes rapid learning of new tasks from the same distribution remains an open problem in meta-learning. Typically, previous works have approached this issue either by attempting to train a neural network that directly produces updates or by attempting to learn better initialisations or scaling factors for a gradient-based update rule. Both of these approaches pose challenges. On one hand, directly producing an update forgoes a useful inductive bias and can easily lead to non-converging behaviour. On the other hand, approaches that try to control a gradient-based update rule typically resort to computing gradients through the learning process to obtain their meta-gradients, leading to methods that can not scale beyond few-shot task adaptation. In this work, we propose Warped Gradient Descent (WarpGrad), a method that intersects these approaches to mitigate their limitations. WarpGrad meta-learns an efficiently parameterised preconditioning matrix that facilitates gradient descent across the task distribution. Preconditioning arises by interleaving non-linear layers, referred to as warp-layers, between the layers of a task-learner. Warp-layers are meta-learned without backpropagating through the task training process in a manner similar to methods that learn to directly produce updates. WarpGrad is computationally efficient, easy to implement, and can scale to arbitrarily large meta-learning problems. We provide a geometrical interpretation of the approach and evaluate its effectiveness in a variety of settings, including few-shot, standard supervised, continual and reinforcement learning.","url_abs":"https://arxiv.org/abs/1909.00025v2","url_pdf":"https://arxiv.org/pdf/1909.00025v2.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":"meta-learning-with-warped-gradient-descent","repo_url":"https://github.com/flennerhag/warpgrad","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"few-shot-learning","task_name":"Few-Shot Learning"},{"task_slug":"inductive-bias","task_name":"Inductive Bias"},{"task_slug":"meta-learning","task_name":"Meta-Learning"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1909.00025","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1909.00025"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/flennerhag/warpgrad","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"unverified":16},"by_repo_kind":{"official":{"samples":16,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"32b845b0bdefaa12","entry":"approx_step","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/warpgrad/warpgrad/utils.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/warpgrad/warpgrad/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"32b845b0bdefaa12"}},{"code_sha256_prefix":"62b6e465c6d6a88f","entry":"build_kwargs","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/omniglot/utils.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/omniglot/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"62b6e465c6d6a88f"}},{"code_sha256_prefix":"e96d229a005e9d51","entry":"compute_ncorrect","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/omniglot/utils.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/omniglot/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e96d229a005e9d51"}},{"code_sha256_prefix":"fc9694f7c4cfc1dc","entry":"convert_arg","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/omniglot/utils.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/omniglot/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"fc9694f7c4cfc1dc"}},{"code_sha256_prefix":"e6117b3cf555d86a","entry":"gen_job","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/omniglot/run_multi.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/omniglot/run_multi.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e6117b3cf555d86a"}},{"code_sha256_prefix":"55501ab84bac6947","entry":"get_async","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/omniglot/run_multi.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/omniglot/run_multi.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"55501ab84bac6947"}},{"code_sha256_prefix":"747bb1ba3659822e","entry":"get_model_config","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/omniglot/run_multi.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/omniglot/run_multi.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"747bb1ba3659822e"}},{"code_sha256_prefix":"9031a2aac8c79c7a","entry":"grad","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/warpgrad/warpgrad/utils.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/warpgrad/warpgrad/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"9031a2aac8c79c7a"}},{"code_sha256_prefix":"96ac1c28c4f7fb11","entry":"initialize","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/maze_navigation/maze.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/maze_navigation/maze.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"96ac1c28c4f7fb11"}},{"code_sha256_prefix":"52200b78221c5ecd","entry":"is_reward","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/maze_navigation/maze.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/maze_navigation/maze.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"52200b78221c5ecd"}},{"code_sha256_prefix":"82e3e23b84f083b9","entry":"line_seg_len","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/warpgrad/warpgrad/updaters.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/warpgrad/warpgrad/updaters.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"82e3e23b84f083b9"}},{"code_sha256_prefix":"b677d9dfa8aa5ea4","entry":"load_args","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/omniglot/monitor.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/omniglot/monitor.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"b677d9dfa8aa5ea4"}},{"code_sha256_prefix":"fe0761d80f06cb62","entry":"load_results","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/omniglot/monitor.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/omniglot/monitor.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"fe0761d80f06cb62"}},{"code_sha256_prefix":"953b2b83b9cf396a","entry":"step","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/warpgrad/warpgrad/utils.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/warpgrad/warpgrad/utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"953b2b83b9cf396a"}},{"code_sha256_prefix":"e3f2074a947300ac","entry":"step_fun","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/maze_navigation/maze.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/maze_navigation/maze.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e3f2074a947300ac"}},{"code_sha256_prefix":"ade47f8be3c0ee8d","entry":"warp_on_same_loss","repo":"flennerhag/warpgrad","repo_kind":"official","path":"src/warpgrad/warpgrad/updaters.py","file_url":"https://github.com/flennerhag/warpgrad/blob/HEAD/src/warpgrad/warpgrad/updaters.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"ade47f8be3c0ee8d"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}