{"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/visualizing-the-loss-landscape-of-neural-nets","title":"Visualizing the Loss Landscape of Neural Nets","arxiv_id":"1712.09913","date":"2017-12-28","proceeding":"ICLR 2018 1","authors":["Hao Li","Zheng Xu","Gavin Taylor","Christoph Studer","Tom Goldstein"],"abstract":"Neural network training relies on our ability to find \"good\" minimizers of\nhighly non-convex loss functions. It is well-known that certain network\narchitecture designs (e.g., skip connections) produce loss functions that train\neasier, and well-chosen training parameters (batch size, learning rate,\noptimizer) produce minimizers that generalize better. However, the reasons for\nthese differences, and their effects on the underlying loss landscape, are not\nwell understood. In this paper, we explore the structure of neural loss\nfunctions, and the effect of loss landscapes on generalization, using a range\nof visualization methods. First, we introduce a simple \"filter normalization\"\nmethod that helps us visualize loss function curvature and make meaningful\nside-by-side comparisons between loss functions. Then, using a variety of\nvisualizations, we explore how network architecture affects the loss landscape,\nand how training parameters affect the shape of minimizers.","url_abs":"http://arxiv.org/abs/1712.09913v3","url_pdf":"http://arxiv.org/pdf/1712.09913v3.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":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/tomgoldstein/loss-landscape","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/JoelNiklaus/loss_landscape","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/QLemma/qLEET","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/StephenThacker/Visualiation-of-Loss-Function","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/activatedgeek/function-space-map","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"jax","reach":null},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/cfellicious/loss-visualization","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"caffe2","reach":{"status":"unanswered"}},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/marcellodebernardi/loss-landscapes","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/mi-peng/sparse-sharpness-aware-minimization","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/okn-yu/Visualizing-the-Loss-Landscape-of-Neural-Nets","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/ramanarao1997/Visualization-of-Loss-Landscapes-in-Neural-Networks","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"visualizing-the-loss-landscape-of-neural-nets","repo_url":"https://github.com/Westlake-AI/openmixup","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1712.09913","atlas_url":"https://app.syntology.ai/?focus=1712.09913","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1712.09913"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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/Westlake-AI/openmixup","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/marcellodebernardi/loss-landscapes","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/activatedgeek/function-space-map","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/StephenThacker/Visualiation-of-Loss-Function","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/okn-yu/Visualizing-the-Loss-Landscape-of-Neural-Nets","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/QLemma/qLEET","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/ramanarao1997/Visualization-of-Loss-Landscapes-in-Neural-Networks","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/tomgoldstein/loss-landscape","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/mi-peng/sparse-sharpness-aware-minimization","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/JoelNiklaus/loss_landscape","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/cfellicious/loss-visualization","reach":{"status":"unanswered"}}],"summary":{"ran":2,"ran_fixture":1,"unverified":5},"by_repo_kind":{"listed":{"samples":8,"ran":3,"repositories":6}},"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":3,"samples":[{"code_sha256_prefix":"de98a260f327adb1","entry":"ModelParameters","repo":"marcellodebernardi/loss-landscapes","repo_kind":"listed","path":"loss_landscapes/model_interface/model_parameters.py","file_url":"https://github.com/marcellodebernardi/loss-landscapes/blob/HEAD/loss_landscapes/model_interface/model_parameters.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"de98a260f327adb1"}},{"code_sha256_prefix":"ea488312c43ec2de","entry":"NormalizeVector","repo":"StephenThacker/Visualiation-of-Loss-Function","repo_kind":"listed","path":"PlotLossFunction.py","file_url":"https://github.com/StephenThacker/Visualiation-of-Loss-Function/blob/HEAD/PlotLossFunction.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ea488312c43ec2de"}},{"code_sha256_prefix":"d50731bae3d21b5a","entry":"perturb_params","repo":"activatedgeek/function-space-map","repo_kind":"listed","path":"experiments/evaluate_landscape.py","file_url":"https://github.com/activatedgeek/function-space-map/blob/HEAD/experiments/evaluate_landscape.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"d50731bae3d21b5a"}},{"code_sha256_prefix":"a58cf2b132df8df9","entry":"filter_normalize","repo":"marcellodebernardi/loss-landscapes","repo_kind":"listed","path":"loss_landscapes/model_interface/model_parameters.py","file_url":"https://github.com/marcellodebernardi/loss-landscapes/blob/HEAD/loss_landscapes/model_interface/model_parameters.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a58cf2b132df8df9"}},{"code_sha256_prefix":"465e856d3ec6e818","entry":"normalize_direction","repo":"Westlake-AI/openmixup","repo_kind":"listed","path":"openmixup/utils/loss_landscape_utils.py","file_url":"https://github.com/Westlake-AI/openmixup/blob/HEAD/openmixup/utils/loss_landscape_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":"465e856d3ec6e818"}},{"code_sha256_prefix":"8fda4d9878d66f73","entry":"normalize_direction","repo":"JoelNiklaus/loss_landscape","repo_kind":"listed","path":"net_plotter.py","file_url":"https://github.com/JoelNiklaus/loss_landscape/blob/HEAD/net_plotter.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"8fda4d9878d66f73"}},{"code_sha256_prefix":"47f610cf6e37def7","entry":"normalize_direction","repo":"okn-yu/Visualizing-the-Loss-Landscape-of-Neural-Nets","repo_kind":"listed","path":"src/directions.py","file_url":"https://github.com/okn-yu/Visualizing-the-Loss-Landscape-of-Neural-Nets/blob/HEAD/src/directions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"47f610cf6e37def7"}},{"code_sha256_prefix":"89b9484af736d895","entry":"normalize_directions_for_weights","repo":"okn-yu/Visualizing-the-Loss-Landscape-of-Neural-Nets","repo_kind":"listed","path":"src/directions.py","file_url":"https://github.com/okn-yu/Visualizing-the-Loss-Landscape-of-Neural-Nets/blob/HEAD/src/directions.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"89b9484af736d895"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}