{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:XNH54SPCN4MF4Z5F5LIWPNGWNM","short_pith_number":"pith:XNH54SPC","schema_version":"1.0","canonical_sha256":"bb4fde49e26f185e67a5ead167b4d66b346285ba18233051d35decd771cac313","source":{"kind":"arxiv","id":"2409.05800","version":1},"attestation_state":"computed","paper":{"title":"Input Space Mode Connectivity in Deep Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cond-mat.stat-mech","cs.CV"],"primary_cat":"cs.LG","authors_text":"David Krueger, Jakub Vrabel, Ori Shem-Ur, Yaron Oz","submitted_at":"2024-09-09T17:03:43Z","abstract_excerpt":"We extend the concept of loss landscape mode connectivity to the input space of deep neural networks. Mode connectivity was originally studied within parameter space, where it describes the existence of low-loss paths between different solutions (loss minimizers) obtained through gradient descent. We present theoretical and empirical evidence of its presence in the input space of deep networks, thereby highlighting the broader nature of the phenomenon. We observe that different input images with similar predictions are generally connected, and for trained models, the path tends to be simple, w"},"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":"2409.05800","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-09-09T17:03:43Z","cross_cats_sorted":["cond-mat.stat-mech","cs.CV"],"title_canon_sha256":"cb91eda8abbcbab68a87af396a34280c0d10189e839831255b18d18d544e463c","abstract_canon_sha256":"b226bfa336994d26a95d5f40acc6352bc5a0f749bb39afd9da0eb2fb5bf5710b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:04:54.318768Z","signature_b64":"+cXJcnC+e1qMQIgaUdga3fhysyS1UA2ZSWA4XouxoAYqQiXDj73h+dMav0ZTfMlAKSJ/MmQVkcXIqhcbqChsDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bb4fde49e26f185e67a5ead167b4d66b346285ba18233051d35decd771cac313","last_reissued_at":"2026-07-05T09:04:54.318356Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:04:54.318356Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Input Space Mode Connectivity in Deep Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cond-mat.stat-mech","cs.CV"],"primary_cat":"cs.LG","authors_text":"David Krueger, Jakub Vrabel, Ori Shem-Ur, Yaron Oz","submitted_at":"2024-09-09T17:03:43Z","abstract_excerpt":"We extend the concept of loss landscape mode connectivity to the input space of deep neural networks. Mode connectivity was originally studied within parameter space, where it describes the existence of low-loss paths between different solutions (loss minimizers) obtained through gradient descent. We present theoretical and empirical evidence of its presence in the input space of deep networks, thereby highlighting the broader nature of the phenomenon. We observe that different input images with similar predictions are generally connected, and for trained models, the path tends to be simple, w"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.05800","kind":"arxiv","version":1},"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/2409.05800/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":"2409.05800","created_at":"2026-07-05T09:04:54.318411+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.05800v1","created_at":"2026-07-05T09:04:54.318411+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.05800","created_at":"2026-07-05T09:04:54.318411+00:00"},{"alias_kind":"pith_short_12","alias_value":"XNH54SPCN4MF","created_at":"2026-07-05T09:04:54.318411+00:00"},{"alias_kind":"pith_short_16","alias_value":"XNH54SPCN4MF4Z5F","created_at":"2026-07-05T09:04:54.318411+00:00"},{"alias_kind":"pith_short_8","alias_value":"XNH54SPC","created_at":"2026-07-05T09:04:54.318411+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.19409","citing_title":"Unlocking the Potential of Continual Model Merging: An ODE Perspective","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19409","citing_title":"Unlocking the Potential of Continual Model Merging: An ODE Perspective","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19409","citing_title":"Unlocking the Potential of Continual Model Merging: An ODE Perspective","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM","json":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM.json","graph_json":"https://pith.science/api/pith-number/XNH54SPCN4MF4Z5F5LIWPNGWNM/graph.json","events_json":"https://pith.science/api/pith-number/XNH54SPCN4MF4Z5F5LIWPNGWNM/events.json","paper":"https://pith.science/paper/XNH54SPC"},"agent_actions":{"view_html":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM","download_json":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM.json","view_paper":"https://pith.science/paper/XNH54SPC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.05800&json=true","fetch_graph":"https://pith.science/api/pith-number/XNH54SPCN4MF4Z5F5LIWPNGWNM/graph.json","fetch_events":"https://pith.science/api/pith-number/XNH54SPCN4MF4Z5F5LIWPNGWNM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM/action/storage_attestation","attest_author":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM/action/author_attestation","sign_citation":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM/action/citation_signature","submit_replication":"https://pith.science/pith/XNH54SPCN4MF4Z5F5LIWPNGWNM/action/replication_record"}},"created_at":"2026-07-05T09:04:54.318411+00:00","updated_at":"2026-07-05T09:04:54.318411+00:00"}