{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:OPK6D324OTXK7OGKHSIEO4TJ5W","short_pith_number":"pith:OPK6D324","schema_version":"1.0","canonical_sha256":"73d5e1ef5c74eeafb8ca3c90477269ed9f6f3d542b5301437d1458368fd8ff6d","source":{"kind":"arxiv","id":"2110.06296","version":2},"attestation_state":"computed","paper":{"title":"The Role of Permutation Invariance in Linear Mode Connectivity of Neural Networks","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Behnam Neyshabur, Hanie Sedghi, Olga Saukh, Rahim Entezari","submitted_at":"2021-10-12T19:28:48Z","abstract_excerpt":"In this paper, we conjecture that if the permutation invariance of neural networks is taken into account, SGD solutions will likely have no barrier in the linear interpolation between them. Although it is a bold conjecture, we show how extensive empirical attempts fall short of refuting it. We further provide a preliminary theoretical result to support our conjecture. Our conjecture has implications for lottery ticket hypothesis, distributed training, and ensemble methods."},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2110.06296","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-12T19:28:48Z","cross_cats_sorted":[],"title_canon_sha256":"fd27a8e9b06d88b07df5a16f2dcd44c2087baa249a6d64f2be939e5fbe5a0b3d","abstract_canon_sha256":"ec7920378df3389431ea8c1a4f1dd9e4caa91edde53a5cbf4fbaa1dc5a0d2b3b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:37:19.255892Z","signature_b64":"qIZmCCWscjKq2ETQ/8hIjdKH4iK0mtvsvrreDUAR+O7/Znf2Rt7IC4KlH4EooCv/5iD1HhhzvHKL8BFiPG7sBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"73d5e1ef5c74eeafb8ca3c90477269ed9f6f3d542b5301437d1458368fd8ff6d","last_reissued_at":"2026-07-05T04:37:19.255287Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:37:19.255287Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Role of Permutation Invariance in Linear Mode Connectivity of Neural Networks","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Behnam Neyshabur, Hanie Sedghi, Olga Saukh, Rahim Entezari","submitted_at":"2021-10-12T19:28:48Z","abstract_excerpt":"In this paper, we conjecture that if the permutation invariance of neural networks is taken into account, SGD solutions will likely have no barrier in the linear interpolation between them. Although it is a bold conjecture, we show how extensive empirical attempts fall short of refuting it. We further provide a preliminary theoretical result to support our conjecture. Our conjecture has implications for lottery ticket hypothesis, distributed training, and ensemble methods."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.06296","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2110.06296/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2110.06296","created_at":"2026-07-05T04:37:19.255352+00:00"},{"alias_kind":"arxiv_version","alias_value":"2110.06296v2","created_at":"2026-07-05T04:37:19.255352+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.06296","created_at":"2026-07-05T04:37:19.255352+00:00"},{"alias_kind":"pith_short_12","alias_value":"OPK6D324OTXK","created_at":"2026-07-05T04:37:19.255352+00:00"},{"alias_kind":"pith_short_16","alias_value":"OPK6D324OTXK7OGK","created_at":"2026-07-05T04:37:19.255352+00:00"},{"alias_kind":"pith_short_8","alias_value":"OPK6D324","created_at":"2026-07-05T04:37:19.255352+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":13,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2607.06628","citing_title":"Cross-Trajectory Chimera Interventions Reveal Dissociable Roles of Weight Magnitude and Direction in Grokking","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2607.08170","citing_title":"Understanding Layer Patching in Model Size Interpolation","ref_index":9,"is_internal_anchor":true},{"citing_arxiv_id":"2606.23607","citing_title":"Scaling Linear Mode Connectivity and Merging to Billion Parameter Pretrained Transformers","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2606.21638","citing_title":"Toward Open Weight Models Without Risks: Separating Public and Private Capabilities in LLMs","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20980","citing_title":"Robusto-2: Benchmarking Humans & VLMs for Autonomous Driving in Lima & New York City","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05538","citing_title":"Less is MoE: Trimming Experts in Domain-Specialist Language Models","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19409","citing_title":"Unlocking the Potential of Continual Model Merging: An ODE Perspective","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24754","citing_title":"Motion-Compensated Weight Compression","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19409","citing_title":"Unlocking the Potential of Continual Model Merging: An ODE Perspective","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19409","citing_title":"Unlocking the Potential of Continual Model Merging: An ODE Perspective","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15459","citing_title":"Don't Stop Me Yet: Sampling Loss Minima via Dissipative Riemannian Mechanics","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2405.07987","citing_title":"The Platonic Representation Hypothesis","ref_index":78,"is_internal_anchor":false},{"citing_arxiv_id":"2212.04089","citing_title":"Editing Models with Task Arithmetic","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W","json":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W.json","graph_json":"https://pith.science/api/pith-number/OPK6D324OTXK7OGKHSIEO4TJ5W/graph.json","events_json":"https://pith.science/api/pith-number/OPK6D324OTXK7OGKHSIEO4TJ5W/events.json","paper":"https://pith.science/paper/OPK6D324"},"agent_actions":{"view_html":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W","download_json":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W.json","view_paper":"https://pith.science/paper/OPK6D324","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2110.06296&json=true","fetch_graph":"https://pith.science/api/pith-number/OPK6D324OTXK7OGKHSIEO4TJ5W/graph.json","fetch_events":"https://pith.science/api/pith-number/OPK6D324OTXK7OGKHSIEO4TJ5W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W/action/storage_attestation","attest_author":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W/action/author_attestation","sign_citation":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W/action/citation_signature","submit_replication":"https://pith.science/pith/OPK6D324OTXK7OGKHSIEO4TJ5W/action/replication_record"}},"created_at":"2026-07-05T04:37:19.255352+00:00","updated_at":"2026-07-05T04:37:19.255352+00:00"}