{"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/a-benchmark-of-selected-algorithmic","title":"A Benchmark of Selected Algorithmic Differentiation Tools on Some Problems in Computer Vision and Machine Learning","arxiv_id":"1807.10129","date":"2018-07-26","proceeding":null,"authors":["Filip Šrajer","Zuzana Kukelova","Andrew Fitzgibbon"],"abstract":"Algorithmic differentiation (AD) allows exact computation of derivatives\ngiven only an implementation of an objective function. Although many AD tools\nare available, a proper and efficient implementation of AD methods is not\nstraightforward. The existing tools are often too different to allow for a\ngeneral test suite. In this paper, we compare fifteen ways of computing\nderivatives including eleven automatic differentiation tools implementing\nvarious methods and written in various languages (C++, F#, MATLAB, Julia and\nPython), two symbolic differentiation tools, finite differences, and\nhand-derived computation.\n  We look at three objective functions from computer vision and machine\nlearning. These objectives are for the most part simple, in the sense that no\niterative loops are involved, and conditional statements are encapsulated in\nfunctions such as {\\tt abs} or {\\tt logsumexp}. However, it is important for\nthe success of algorithmic differentiation that such `simple' objective\nfunctions are handled efficiently, as so many problems in computer vision and\nmachine learning are of this form.\n  Of course, our results depend on programmer skill, and familiarity with the\ntools. However, we contend that this paper presents an important datapoint: a\nskilled programmer devoting roughly a week to each tool produced the timings we\npresent. We have made our implementations available as open source to allow the\ncommunity to replicate and update these benchmarks.","url_abs":"http://arxiv.org/abs/1807.10129v1","url_pdf":"http://arxiv.org/pdf/1807.10129v1.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":"a-benchmark-of-selected-algorithmic","repo_url":"https://github.com/awf/autodiff2","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"a-benchmark-of-selected-algorithmic","repo_url":"https://github.com/awf/ADBench","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"a-benchmark-of-selected-algorithmic","repo_url":"https://github.com/microsoft/ADBench","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"machine-learning","task_name":"BIG-bench Machine Learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1807.10129","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1807.10129"}},"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/microsoft/ADBench","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/awf/ADBench","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/awf/autodiff2","reach":{"status":"ok"}}],"summary":{"ran_draft_wrong":1,"unverified":12},"by_repo_kind":{"listed":{"samples":13,"ran":1,"repositories":2}},"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":12,"samples":[{"code_sha256_prefix":"74c9da4f6b8b0da0","entry":"objective_display_name","repo":"microsoft/ADBench","repo_kind":"listed","path":"ADBench/plot_graphs.py","file_url":"https://github.com/microsoft/ADBench/blob/HEAD/ADBench/plot_graphs.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"74c9da4f6b8b0da0"}},{"code_sha256_prefix":"9cefd234c81926f6","entry":"cap_str","repo":"awf/ADBench","repo_kind":"listed","path":"ADBench/utils.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/ADBench/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"9cefd234c81926f6"}},{"code_sha256_prefix":"475acdd13fbe59f5","entry":"filepath_to_basename","repo":"awf/ADBench","repo_kind":"listed","path":"src/python/runner/Filepaths.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/src/python/runner/Filepaths.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"475acdd13fbe59f5"}},{"code_sha256_prefix":"1b9349bc26fa72ce","entry":"filepath_to_dirname","repo":"awf/ADBench","repo_kind":"listed","path":"src/python/runner/Filepaths.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/src/python/runner/Filepaths.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"1b9349bc26fa72ce"}},{"code_sha256_prefix":"4ced703c7a310071","entry":"find_repeats_for_minimum_measurable_time","repo":"awf/ADBench","repo_kind":"listed","path":"src/python/runner/Benchmark.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/src/python/runner/Benchmark.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"4ced703c7a310071"}},{"code_sha256_prefix":"acc7f1787dc94b66","entry":"get_fn","repo":"awf/ADBench","repo_kind":"listed","path":"ADBench/utils.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/ADBench/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"acc7f1787dc94b66"}},{"code_sha256_prefix":"1151cb2d78194e5b","entry":"get_tool","repo":"awf/ADBench","repo_kind":"listed","path":"ADBench/utils.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/ADBench/utils.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"1151cb2d78194e5b"}},{"code_sha256_prefix":"e9ac370154cec5f8","entry":"load_matrix","repo":"awf/ADBench","repo_kind":"listed","path":"ADBench/check_J.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/ADBench/check_J.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"e9ac370154cec5f8"}},{"code_sha256_prefix":"3fdcf207a614916b","entry":"measure_shortest_time","repo":"awf/ADBench","repo_kind":"listed","path":"src/python/runner/Benchmark.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/src/python/runner/Benchmark.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"3fdcf207a614916b"}},{"code_sha256_prefix":"4f29a61bb2a73796","entry":"module_load","repo":"awf/ADBench","repo_kind":"listed","path":"src/python/runner/ModuleLoader.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/src/python/runner/ModuleLoader.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"4f29a61bb2a73796"}},{"code_sha256_prefix":"3bb2c5a077ed30bc","entry":"modulepath_to_basename","repo":"awf/ADBench","repo_kind":"listed","path":"src/python/runner/Filepaths.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/src/python/runner/Filepaths.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"3bb2c5a077ed30bc"}},{"code_sha256_prefix":"08deada56461fd30","entry":"round_sf","repo":"awf/ADBench","repo_kind":"listed","path":"ADBench/check_J.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/ADBench/check_J.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"08deada56461fd30"}},{"code_sha256_prefix":"6696b6fe355729ef","entry":"test_equality","repo":"awf/ADBench","repo_kind":"listed","path":"ADBench/check_J.py","file_url":"https://github.com/awf/ADBench/blob/HEAD/ADBench/check_J.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":false,"mcp_get_code":{"code_sha256":"6696b6fe355729ef"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}