{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:DXGMSHC2QWHQZ7R7F5Z6BIABDP","short_pith_number":"pith:DXGMSHC2","schema_version":"1.0","canonical_sha256":"1dccc91c5a858f0cfe3f2f73e0a0011bc885e845013061d13414d4245fd33d19","source":{"kind":"arxiv","id":"2605.20610","version":1},"attestation_state":"computed","paper":{"title":"Beyond Routing: Characterising Expert Tuning and Representation in Vision Mixture-of-Experts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Gene Tangtartharakul, Katherine R. Storrs","submitted_at":"2026-05-20T01:55:40Z","abstract_excerpt":"Mixture-of-Experts (MoE) models are often interpreted by analysing which categories are routed to which experts. However, routing alone does not reveal what each expert actually encodes. We train sparsely-gated convolutional MoE models with a contrastive objective on natural images and characterise expert specialisation using tools from visual neuroscience. Extending from gating-level to expert-level analyses, we measure per-expert category separability, and per-expert tuning using the most exciting inputs. Extending from category-level to feature-level explanations, we interpret tuning via se"},"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":"2605.20610","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-20T01:55:40Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0e4342cd812ff966924d1476e55f6ac60a1aee1322b61e0c0ecb6bec37720f32","abstract_canon_sha256":"d03c9ed66d2feecbafdfb7a75554595fdfe270ad74508c742f27344fdcb9718e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:04:44.471950Z","signature_b64":"AbBG0JH4YgWjYQ33W86Gs+wpLXBjDLJ2rCdSYpfprRTeyCq8wZhQ7A6ci+Ji9L3UEpIWAhvR6t1f9116/gEgCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1dccc91c5a858f0cfe3f2f73e0a0011bc885e845013061d13414d4245fd33d19","last_reissued_at":"2026-05-21T01:04:44.471296Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:04:44.471296Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Beyond Routing: Characterising Expert Tuning and Representation in Vision Mixture-of-Experts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Gene Tangtartharakul, Katherine R. Storrs","submitted_at":"2026-05-20T01:55:40Z","abstract_excerpt":"Mixture-of-Experts (MoE) models are often interpreted by analysing which categories are routed to which experts. However, routing alone does not reveal what each expert actually encodes. We train sparsely-gated convolutional MoE models with a contrastive objective on natural images and characterise expert specialisation using tools from visual neuroscience. Extending from gating-level to expert-level analyses, we measure per-expert category separability, and per-expert tuning using the most exciting inputs. Extending from category-level to feature-level explanations, we interpret tuning via se"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20610","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/2605.20610/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":"2605.20610","created_at":"2026-05-21T01:04:44.471412+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.20610v1","created_at":"2026-05-21T01:04:44.471412+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20610","created_at":"2026-05-21T01:04:44.471412+00:00"},{"alias_kind":"pith_short_12","alias_value":"DXGMSHC2QWHQ","created_at":"2026-05-21T01:04:44.471412+00:00"},{"alias_kind":"pith_short_16","alias_value":"DXGMSHC2QWHQZ7R7","created_at":"2026-05-21T01:04:44.471412+00:00"},{"alias_kind":"pith_short_8","alias_value":"DXGMSHC2","created_at":"2026-05-21T01:04:44.471412+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP","json":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP.json","graph_json":"https://pith.science/api/pith-number/DXGMSHC2QWHQZ7R7F5Z6BIABDP/graph.json","events_json":"https://pith.science/api/pith-number/DXGMSHC2QWHQZ7R7F5Z6BIABDP/events.json","paper":"https://pith.science/paper/DXGMSHC2"},"agent_actions":{"view_html":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP","download_json":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP.json","view_paper":"https://pith.science/paper/DXGMSHC2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.20610&json=true","fetch_graph":"https://pith.science/api/pith-number/DXGMSHC2QWHQZ7R7F5Z6BIABDP/graph.json","fetch_events":"https://pith.science/api/pith-number/DXGMSHC2QWHQZ7R7F5Z6BIABDP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP/action/storage_attestation","attest_author":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP/action/author_attestation","sign_citation":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP/action/citation_signature","submit_replication":"https://pith.science/pith/DXGMSHC2QWHQZ7R7F5Z6BIABDP/action/replication_record"}},"created_at":"2026-05-21T01:04:44.471412+00:00","updated_at":"2026-05-21T01:04:44.471412+00:00"}