{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:KELUQXQWF7DKR3L65DC7PLVFUT","short_pith_number":"pith:KELUQXQW","schema_version":"1.0","canonical_sha256":"5117485e162fc6a8ed7ee8c5f7aea5a4e2d5d5156b98be3aa52fefc450741f96","source":{"kind":"arxiv","id":"2605.30992","version":1},"attestation_state":"computed","paper":{"title":"Eigenvectors of Experts are Training-free Non-collapsing Routers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Giang Do, Hung Le, Truyen Tran","submitted_at":"2026-05-29T08:27:10Z","abstract_excerpt":"Sparse Mixture of Experts (SMoE) architectures improve the training efficiency of Large Language Models (LLMs) by routing input tokens to a selected subset of specialized experts. Despite their remarkable success, both training and inference in SMoE models suffer from the expert collapse issue (Chi et al., 2022), which degrades model performance. Prior studies primarily focus on improving the router; however, such methods rely on training from scratch or fine-tuning, which requires high computational and data-processing costs. Furthermore, we demonstrate that, despite these efforts, the issue "},"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":"2605.30992","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T08:27:10Z","cross_cats_sorted":[],"title_canon_sha256":"c2072ea808ad0e3d8eb2e8da6433b43ee737ff4ae0cfa74985d433e9aa434f1d","abstract_canon_sha256":"2db1ae281539c245c530faa360aa931996d27b88cd44725e8ebaee05857b6338"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-01T01:03:29.208283Z","signature_b64":"TRTeDKb88aAlvvbMhuZALiqjck10tN1M2BTcku0mWV4rvuDb4OYA8JmieYzxjOe3BTz1kJ3MS8UaiS7FQgEoAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5117485e162fc6a8ed7ee8c5f7aea5a4e2d5d5156b98be3aa52fefc450741f96","last_reissued_at":"2026-06-01T01:03:29.207670Z","signature_status":"signed_v1","first_computed_at":"2026-06-01T01:03:29.207670Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Eigenvectors of Experts are Training-free Non-collapsing Routers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Giang Do, Hung Le, Truyen Tran","submitted_at":"2026-05-29T08:27:10Z","abstract_excerpt":"Sparse Mixture of Experts (SMoE) architectures improve the training efficiency of Large Language Models (LLMs) by routing input tokens to a selected subset of specialized experts. Despite their remarkable success, both training and inference in SMoE models suffer from the expert collapse issue (Chi et al., 2022), which degrades model performance. Prior studies primarily focus on improving the router; however, such methods rely on training from scratch or fine-tuning, which requires high computational and data-processing costs. Furthermore, we demonstrate that, despite these efforts, the issue "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.30992","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/2605.30992/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":"2605.30992","created_at":"2026-06-01T01:03:29.207775+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.30992v1","created_at":"2026-06-01T01:03:29.207775+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.30992","created_at":"2026-06-01T01:03:29.207775+00:00"},{"alias_kind":"pith_short_12","alias_value":"KELUQXQWF7DK","created_at":"2026-06-01T01:03:29.207775+00:00"},{"alias_kind":"pith_short_16","alias_value":"KELUQXQWF7DKR3L6","created_at":"2026-06-01T01:03:29.207775+00:00"},{"alias_kind":"pith_short_8","alias_value":"KELUQXQW","created_at":"2026-06-01T01:03:29.207775+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/KELUQXQWF7DKR3L65DC7PLVFUT","json":"https://pith.science/pith/KELUQXQWF7DKR3L65DC7PLVFUT.json","graph_json":"https://pith.science/api/pith-number/KELUQXQWF7DKR3L65DC7PLVFUT/graph.json","events_json":"https://pith.science/api/pith-number/KELUQXQWF7DKR3L65DC7PLVFUT/events.json","paper":"https://pith.science/paper/KELUQXQW"},"agent_actions":{"view_html":"https://pith.science/pith/KELUQXQWF7DKR3L65DC7PLVFUT","download_json":"https://pith.science/pith/KELUQXQWF7DKR3L65DC7PLVFUT.json","view_paper":"https://pith.science/paper/KELUQXQW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.30992&json=true","fetch_graph":"https://pith.science/api/pith-number/KELUQXQWF7DKR3L65DC7PLVFUT/graph.json","fetch_events":"https://pith.science/api/pith-number/KELUQXQWF7DKR3L65DC7PLVFUT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KELUQXQWF7DKR3L65DC7PLVFUT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KELUQXQWF7DKR3L65DC7PLVFUT/action/storage_attestation","attest_author":"https://pith.science/pith/KELUQXQWF7DKR3L65DC7PLVFUT/action/author_attestation","sign_citation":"https://pith.science/pith/KELUQXQWF7DKR3L65DC7PLVFUT/action/citation_signature","submit_replication":"https://pith.science/pith/KELUQXQWF7DKR3L65DC7PLVFUT/action/replication_record"}},"created_at":"2026-06-01T01:03:29.207775+00:00","updated_at":"2026-06-01T01:03:29.207775+00:00"}