{"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/arxiv-2605-10047","title":"Rethinking Loss Reweighting for Imbalance Learning as an Inverse Problem: A Neural Collapse Point of View","arxiv_id":"2605.10047","date":"2026-05-11","proceeding":"ICML","authors":["Jinping Wang","Zixin Tong","Zhiwu Xie","Zhiqiang Gao"],"abstract":"Loss reweighting is a widely used strategy for long-tailed classification, but existing reweighting strategies often rely on heuristics and rarely define a well-specified target. Inspired by Neural Collapse (NC), the ideal simplex Equiangular Tight Frame (ETF) terminal geometry suggests equal per-class average loss as a reasonable target for reweighting. Based on the ideal equal loss objective, we consider loss reweighting as an inverse problem and propose an inverse-view reweighting strategy that infers class weights dynamically to match this ideal objective. Empirically, NC metrics suggest our method can effectively reduce the loss imbalance coefficient and closer alignment with NC geometry while consistently outperforming strong long-tailed baselines on different datasets.","url_abs":"https://arxiv.org/abs/2605.10047","url_pdf":"https://arxiv.org/pdf/2605.10047","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2605.10047","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2605.10047"}},"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":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting","reach":null}],"summary":{"ran":4,"ran_draft_wrong":2,"unverified":6},"by_repo_kind":{"found_in_text":{"samples":12,"ran":6,"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":12,"samples":[{"code_sha256_prefix":"c23d34000bafe4c6","entry":"cifar_partition_by_grad_nogroup_bcl","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"utils.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"c23d34000bafe4c6"}},{"code_sha256_prefix":"2a80220dabcb742a","entry":"conv1x1","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"models/resnext.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/models/resnext.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"2a80220dabcb742a"}},{"code_sha256_prefix":"600ff2c45e0de056","entry":"conv3x3","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"models/resnext.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/models/resnext.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"600ff2c45e0de056"}},{"code_sha256_prefix":"9c12115c283ca5af","entry":"gradient_normalizers","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"min_norm_solvers.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/min_norm_solvers.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"9c12115c283ca5af"}},{"code_sha256_prefix":"3612da0362c505e5","entry":"shot_acc","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"utils.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"3612da0362c505e5"}},{"code_sha256_prefix":"a4b471197204b414","entry":"translate_x_rel","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"randaugment.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/randaugment.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"a4b471197204b414"}},{"code_sha256_prefix":"0a7500eb917e573a","entry":"imagenet_partition_by_grad_nogroup_bcl","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"utils.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/utils.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":"0a7500eb917e573a"}},{"code_sha256_prefix":"a46153beb2cbece1","entry":"resnet32","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"dataset/resnet_cifar.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/dataset/resnet_cifar.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":"a46153beb2cbece1"}},{"code_sha256_prefix":"a8d9ab2416197501","entry":"resnet32","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"models/resnet_cifar.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/models/resnet_cifar.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":"a8d9ab2416197501"}},{"code_sha256_prefix":"7be5050a9402a4c8","entry":"resnet50","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"models/resnext.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/models/resnext.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":"7be5050a9402a4c8"}},{"code_sha256_prefix":"7b4e3d0bd192b677","entry":"shear_x","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"randaugment.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/randaugment.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":"7b4e3d0bd192b677"}},{"code_sha256_prefix":"e31c4b500a240e69","entry":"shear_y","repo":"tongzixin716716/Inverse-Loss-Reweighting","repo_kind":"found_in_text","path":"randaugment.py","file_url":"https://github.com/tongzixin716716/Inverse-Loss-Reweighting/blob/HEAD/randaugment.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":"e31c4b500a240e69"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}