{"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-2603-06248","title":"Gradient Flow Polarizes Softmax Outputs towards Low-Entropy Solutions","arxiv_id":"2603.06248","date":"2026-03-06","proceeding":null,"authors":["Aditya Varre","Mark Rofin","Nicolas Flammarion"],"abstract":"Understanding the intricate non-convex training dynamics of softmax-based models is crucial for explaining the empirical success of transformers. In this article, we analyze the gradient flow dynamics of the value-softmax model, defined as ${L}(\\mathbf{V} σ(\\mathbf{a}))$, where $\\mathbf{V}$ and $\\mathbf{a}$ are a learnable value matrix and attention vector, respectively. As the matrix times softmax vector parameterization constitutes the core building block of self-attention, our analysis provides direct insight into transformer's training dynamics. We reveal that gradient flow on this structure inherently drives the optimization toward solutions characterized by low-entropy outputs. We demonstrate the universality of this polarizing effect across various objectives, including logistic and square loss. Furthermore, we discuss the practical implications of these theoretical results, offering a formal mechanism for empirical phenomena such as attention sinks and massive activations.","url_abs":"https://arxiv.org/abs/2603.06248","url_pdf":"https://arxiv.org/pdf/2603.06248","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":{"syntology_url":"https://syntology.ai/paper/2603.06248","atlas_url":"https://app.syntology.ai/?focus=2603.06248","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2603.06248"}},"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":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/tml-epfl/softmax","reach":null}],"summary":{"unverified":10},"by_repo_kind":{"found_in_text":{"samples":10,"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":10,"samples":[{"code_sha256_prefix":"de1a6b54bb6d7eb2","entry":"evaluate","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"exp_classification.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/exp_classification.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":"de1a6b54bb6d7eb2"}},{"code_sha256_prefix":"195ef9d2094e5fa7","entry":"evaluate","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"exp_induction.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/exp_induction.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":"195ef9d2094e5fa7"}},{"code_sha256_prefix":"dbc815742c417081","entry":"get_metrics_for_batch","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"exp_classification.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/exp_classification.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":"dbc815742c417081"}},{"code_sha256_prefix":"179263c917ddf20b","entry":"get_metrics_for_batch","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"exp_induction.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/exp_induction.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":"179263c917ddf20b"}},{"code_sha256_prefix":"c50658a8f12d5f40","entry":"get_metrics_for_dataset","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"exp_induction.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/exp_induction.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":"c50658a8f12d5f40"}},{"code_sha256_prefix":"a91bbe5c96744514","entry":"linear_attention","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"src/attention.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/src/attention.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":"a91bbe5c96744514"}},{"code_sha256_prefix":"dea3bf88d257b9b2","entry":"print_induction_sequence","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"src/datasets.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/src/datasets.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":"dea3bf88d257b9b2"}},{"code_sha256_prefix":"05229e47182eaaf7","entry":"sigmoid_attention","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"src/attention.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/src/attention.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":"05229e47182eaaf7"}},{"code_sha256_prefix":"cac10aee0417a8c9","entry":"softmax_attention","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"src/attention.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/src/attention.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":"cac10aee0417a8c9"}},{"code_sha256_prefix":"0dbeea554881e4bd","entry":"train","repo":"tml-epfl/softmax","repo_kind":"found_in_text","path":"exp_classification.py","file_url":"https://github.com/tml-epfl/softmax/blob/HEAD/exp_classification.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":"0dbeea554881e4bd"}}]},"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":{"kind":"leaderboard_placements","source":"Syntology's leaderboard-shaped extractor over the paper's own arXiv-HTML tables: a model pointed at a cell, the number was read from that cell and checked against the board's metric, dataset, split and scale, and an independent check accepted the entry; not reviewed by the paper's authors or the archive's editors","extractor_model":"global.anthropic.claude-sonnet-4-5-20250929-v1:0","verifier_model":null,"prompt_sha":"fa63d4bb9d755694","coverage":{"sentence":"Syntology has checked 7,081 of the 9,623 papers on this site that are newer than the archive; results from the others appear after they are checked.","papers_newer_than_archive":9623,"papers_checked":7081},"entries":[],"not_placed":{"boards":0,"rejected_by_independent_check":0,"refused_by_a_rule":0,"check_did_not_answer":0,"proposed_without_a_cell":0,"declined_by_site":0}}}