{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:J4OKRBCMM37B6IL6DY2DV2JOQZ","short_pith_number":"pith:J4OKRBCM","schema_version":"1.0","canonical_sha256":"4f1ca8844c66fe1f217e1e343ae92e864945499348639dda4769ecee70ad7775","source":{"kind":"arxiv","id":"2605.24270","version":1},"attestation_state":"computed","paper":{"title":"Safety-Oriented Routing Analysis of Mixtral MoE Under Benign and Harmful Prompts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.AI","authors_text":"Md Nurul Absar Siddiky","submitted_at":"2026-05-22T22:42:23Z","abstract_excerpt":"Sparse mixture-of-experts (MoE) language models activate only a small subset of parameters for each token, making router behavior a central part of model computation. This paper studies routing behavior of Mixtral 8x7B-Instruct under benign and harmful prompts using two complementary signals: activation-based routing scores derived from expert selection frequencies and gradient-based scores derived from router-gate sensitivities. We analyze expert- and layer-level routing behavior and conduct expert-suppression interventions. The results show that activation-based expert usage is broad and lon"},"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":"2605.24270","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-22T22:42:23Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"8cdf099d67f862f05fc4f40754385deb7ff7547f6a2cce80df380d07311b5c08","abstract_canon_sha256":"380f51fe21972933d02d618a7fb9a6c6466e84cddb4bfa718ef18557b6ae5cc7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:02:56.172840Z","signature_b64":"8FBAfb/MEvaVqpY12sOug4pzX0N7baUZzqKVFlrBebVS++FfnmqOzflF/1v7JzT0D6oIjgfF2zbGHZLuETvACw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4f1ca8844c66fe1f217e1e343ae92e864945499348639dda4769ecee70ad7775","last_reissued_at":"2026-05-26T01:02:56.172065Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:02:56.172065Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Safety-Oriented Routing Analysis of Mixtral MoE Under Benign and Harmful Prompts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.AI","authors_text":"Md Nurul Absar Siddiky","submitted_at":"2026-05-22T22:42:23Z","abstract_excerpt":"Sparse mixture-of-experts (MoE) language models activate only a small subset of parameters for each token, making router behavior a central part of model computation. This paper studies routing behavior of Mixtral 8x7B-Instruct under benign and harmful prompts using two complementary signals: activation-based routing scores derived from expert selection frequencies and gradient-based scores derived from router-gate sensitivities. We analyze expert- and layer-level routing behavior and conduct expert-suppression interventions. The results show that activation-based expert usage is broad and lon"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.24270","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/2605.24270/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":"2605.24270","created_at":"2026-05-26T01:02:56.172192+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.24270v1","created_at":"2026-05-26T01:02:56.172192+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.24270","created_at":"2026-05-26T01:02:56.172192+00:00"},{"alias_kind":"pith_short_12","alias_value":"J4OKRBCMM37B","created_at":"2026-05-26T01:02:56.172192+00:00"},{"alias_kind":"pith_short_16","alias_value":"J4OKRBCMM37B6IL6","created_at":"2026-05-26T01:02:56.172192+00:00"},{"alias_kind":"pith_short_8","alias_value":"J4OKRBCM","created_at":"2026-05-26T01:02:56.172192+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/J4OKRBCMM37B6IL6DY2DV2JOQZ","json":"https://pith.science/pith/J4OKRBCMM37B6IL6DY2DV2JOQZ.json","graph_json":"https://pith.science/api/pith-number/J4OKRBCMM37B6IL6DY2DV2JOQZ/graph.json","events_json":"https://pith.science/api/pith-number/J4OKRBCMM37B6IL6DY2DV2JOQZ/events.json","paper":"https://pith.science/paper/J4OKRBCM"},"agent_actions":{"view_html":"https://pith.science/pith/J4OKRBCMM37B6IL6DY2DV2JOQZ","download_json":"https://pith.science/pith/J4OKRBCMM37B6IL6DY2DV2JOQZ.json","view_paper":"https://pith.science/paper/J4OKRBCM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.24270&json=true","fetch_graph":"https://pith.science/api/pith-number/J4OKRBCMM37B6IL6DY2DV2JOQZ/graph.json","fetch_events":"https://pith.science/api/pith-number/J4OKRBCMM37B6IL6DY2DV2JOQZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/J4OKRBCMM37B6IL6DY2DV2JOQZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/J4OKRBCMM37B6IL6DY2DV2JOQZ/action/storage_attestation","attest_author":"https://pith.science/pith/J4OKRBCMM37B6IL6DY2DV2JOQZ/action/author_attestation","sign_citation":"https://pith.science/pith/J4OKRBCMM37B6IL6DY2DV2JOQZ/action/citation_signature","submit_replication":"https://pith.science/pith/J4OKRBCMM37B6IL6DY2DV2JOQZ/action/replication_record"}},"created_at":"2026-05-26T01:02:56.172192+00:00","updated_at":"2026-05-26T01:02:56.172192+00:00"}