{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:ERXKAGEMFI6G3YCSVZKF4XIYFC","short_pith_number":"pith:ERXKAGEM","canonical_record":{"source":{"id":"2409.12136","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-18T17:00:20Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"955f0b56af98400f2e37c0831add31cc7daebefa0a8c388810361f611d501159","abstract_canon_sha256":"dec502203bc14a5ef6e8af242c2e336dd098dfc2ecb2a56586223ab4655c4673"},"schema_version":"1.0"},"canonical_sha256":"246ea0188c2a3c6de052ae545e5d1828928228446735c5be3a2d938a91b2d95f","source":{"kind":"arxiv","id":"2409.12136","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.12136","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"arxiv_version","alias_value":"2409.12136v1","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.12136","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"pith_short_12","alias_value":"ERXKAGEMFI6G","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"pith_short_16","alias_value":"ERXKAGEMFI6G3YCS","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"pith_short_8","alias_value":"ERXKAGEM","created_at":"2026-07-05T09:08:43Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:ERXKAGEMFI6G3YCSVZKF4XIYFC","target":"record","payload":{"canonical_record":{"source":{"id":"2409.12136","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-18T17:00:20Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"955f0b56af98400f2e37c0831add31cc7daebefa0a8c388810361f611d501159","abstract_canon_sha256":"dec502203bc14a5ef6e8af242c2e336dd098dfc2ecb2a56586223ab4655c4673"},"schema_version":"1.0"},"canonical_sha256":"246ea0188c2a3c6de052ae545e5d1828928228446735c5be3a2d938a91b2d95f","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:08:43.159671Z","signature_b64":"iyl+mFJo26Go8UxtWAkZuzXD5A0I/pj8QDqRCMr5RVTzZ1S1dseTquHx1SoS5jDnHiJltd9HBUJ3/s68XHiBAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"246ea0188c2a3c6de052ae545e5d1828928228446735c5be3a2d938a91b2d95f","last_reissued_at":"2026-07-05T09:08:43.159128Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:08:43.159128Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2409.12136","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:08:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"odhKBcHDXxkB3sCBa3VAviT0srBib6RJPzAxkEiGp6KgMeybKNs7zPuc14HledNuTM2sGA/hr2TAU7w81hWwCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T05:31:20.066895Z"},"content_sha256":"d7e08a3f5570e6b5e4ec2a412691f7039c8e7b5a44b1f6c0480915b9669a95e5","schema_version":"1.0","event_id":"sha256:d7e08a3f5570e6b5e4ec2a412691f7039c8e7b5a44b1f6c0480915b9669a95e5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:ERXKAGEMFI6G3YCSVZKF4XIYFC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"GRIN: GRadient-INformed MoE","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Chen Liang, Chenruidong Zhang, Hany Awadalla, Hao Cheng, Jianfeng Gao, Jilong Xue, Liyuan Liu, Masahiro Tanaka, Shuohang Wang, Vishrav Chaudhary, Weizhu Chen, Wenxiang Hu, XiaoDong Liu, Xiaoxia Wu, Yelong Shen, Young Jin Kim, Zeqi Lin","submitted_at":"2024-09-18T17:00:20Z","abstract_excerpt":"Mixture-of-Experts (MoE) models scale more effectively than dense models due to sparse computation through expert routing, selectively activating only a small subset of expert modules. However, sparse computation challenges traditional training practices, as discrete expert routing hinders standard backpropagation and thus gradient-based optimization, which are the cornerstone of deep learning. To better pursue the scaling power of MoE, we introduce GRIN (GRadient-INformed MoE training), which incorporates sparse gradient estimation for expert routing and configures model parallelism to avoid "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.12136","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/2409.12136/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T09:08:43Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"R1NjrwfdaZiZ1GtWuTaQw4yadYcl7d1OVbn0MgRPPCGl4qvIelxl0jew64Y8QP8zKvH/1ihCGeFd3Y1ciKRNCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T05:31:20.067492Z"},"content_sha256":"1ecb22cd73dc4e498523b9f3fa57a134282a1f5f5b23716827d182c3604e8b2e","schema_version":"1.0","event_id":"sha256:1ecb22cd73dc4e498523b9f3fa57a134282a1f5f5b23716827d182c3604e8b2e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ERXKAGEMFI6G3YCSVZKF4XIYFC/bundle.json","state_url":"https://pith.science/pith/ERXKAGEMFI6G3YCSVZKF4XIYFC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ERXKAGEMFI6G3YCSVZKF4XIYFC/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-09T05:31:20Z","links":{"resolver":"https://pith.science/pith/ERXKAGEMFI6G3YCSVZKF4XIYFC","bundle":"https://pith.science/pith/ERXKAGEMFI6G3YCSVZKF4XIYFC/bundle.json","state":"https://pith.science/pith/ERXKAGEMFI6G3YCSVZKF4XIYFC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ERXKAGEMFI6G3YCSVZKF4XIYFC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:ERXKAGEMFI6G3YCSVZKF4XIYFC","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"dec502203bc14a5ef6e8af242c2e336dd098dfc2ecb2a56586223ab4655c4673","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-18T17:00:20Z","title_canon_sha256":"955f0b56af98400f2e37c0831add31cc7daebefa0a8c388810361f611d501159"},"schema_version":"1.0","source":{"id":"2409.12136","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2409.12136","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"arxiv_version","alias_value":"2409.12136v1","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.12136","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"pith_short_12","alias_value":"ERXKAGEMFI6G","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"pith_short_16","alias_value":"ERXKAGEMFI6G3YCS","created_at":"2026-07-05T09:08:43Z"},{"alias_kind":"pith_short_8","alias_value":"ERXKAGEM","created_at":"2026-07-05T09:08:43Z"}],"graph_snapshots":[{"event_id":"sha256:1ecb22cd73dc4e498523b9f3fa57a134282a1f5f5b23716827d182c3604e8b2e","target":"graph","created_at":"2026-07-05T09:08:43Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2409.12136/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Mixture-of-Experts (MoE) models scale more effectively than dense models due to sparse computation through expert routing, selectively activating only a small subset of expert modules. However, sparse computation challenges traditional training practices, as discrete expert routing hinders standard backpropagation and thus gradient-based optimization, which are the cornerstone of deep learning. To better pursue the scaling power of MoE, we introduce GRIN (GRadient-INformed MoE training), which incorporates sparse gradient estimation for expert routing and configures model parallelism to avoid ","authors_text":"Chen Liang, Chenruidong Zhang, Hany Awadalla, Hao Cheng, Jianfeng Gao, Jilong Xue, Liyuan Liu, Masahiro Tanaka, Shuohang Wang, Vishrav Chaudhary, Weizhu Chen, Wenxiang Hu, XiaoDong Liu, Xiaoxia Wu, Yelong Shen, Young Jin Kim, Zeqi Lin","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-18T17:00:20Z","title":"GRIN: GRadient-INformed MoE"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.12136","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d7e08a3f5570e6b5e4ec2a412691f7039c8e7b5a44b1f6c0480915b9669a95e5","target":"record","created_at":"2026-07-05T09:08:43Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"dec502203bc14a5ef6e8af242c2e336dd098dfc2ecb2a56586223ab4655c4673","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-18T17:00:20Z","title_canon_sha256":"955f0b56af98400f2e37c0831add31cc7daebefa0a8c388810361f611d501159"},"schema_version":"1.0","source":{"id":"2409.12136","kind":"arxiv","version":1}},"canonical_sha256":"246ea0188c2a3c6de052ae545e5d1828928228446735c5be3a2d938a91b2d95f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"246ea0188c2a3c6de052ae545e5d1828928228446735c5be3a2d938a91b2d95f","first_computed_at":"2026-07-05T09:08:43.159128Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:08:43.159128Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"iyl+mFJo26Go8UxtWAkZuzXD5A0I/pj8QDqRCMr5RVTzZ1S1dseTquHx1SoS5jDnHiJltd9HBUJ3/s68XHiBAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T09:08:43.159671Z","signed_message":"canonical_sha256_bytes"},"source_id":"2409.12136","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d7e08a3f5570e6b5e4ec2a412691f7039c8e7b5a44b1f6c0480915b9669a95e5","sha256:1ecb22cd73dc4e498523b9f3fa57a134282a1f5f5b23716827d182c3604e8b2e"],"state_sha256":"8c22937f358de90c794c7404d380530b9f133fde731d55b11628a97133e9a8ae"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bqbZYIPN5PXxVfaL2lGyG6olvou5Bs9tAkUctyUo6C8U+23vVTD4pr92RNVCqb0P/SW9BPttAeAGT0cH4CGsCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T05:31:20.072535Z","bundle_sha256":"3d7ab6334aadb5056e855e69bd2b4a0a75cc6e8ab07cd53e9b523110ec7b3832"}}