{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:L5AXDG3PU6VOFEAQEQMRIYOY6F","short_pith_number":"pith:L5AXDG3P","canonical_record":{"source":{"id":"2406.14088","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2024-06-20T08:04:07Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"da7ec43cbeb19100f4ef4804d02de73984a8aa2b0ed119a868b59912cfe19823","abstract_canon_sha256":"11bc9b4cca7caa5a4acc3c936471da2ddd258f74bd4c03cf64a7d1c975954f95"},"schema_version":"1.0"},"canonical_sha256":"5f41719b6fa7aae2901024191461d8f1460affcaef6722aa3a39064d71843e76","source":{"kind":"arxiv","id":"2406.14088","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.14088","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"arxiv_version","alias_value":"2406.14088v2","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.14088","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"pith_short_12","alias_value":"L5AXDG3PU6VO","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"pith_short_16","alias_value":"L5AXDG3PU6VOFEAQ","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"pith_short_8","alias_value":"L5AXDG3P","created_at":"2026-07-05T10:53:10Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:L5AXDG3PU6VOFEAQEQMRIYOY6F","target":"record","payload":{"canonical_record":{"source":{"id":"2406.14088","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2024-06-20T08:04:07Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"da7ec43cbeb19100f4ef4804d02de73984a8aa2b0ed119a868b59912cfe19823","abstract_canon_sha256":"11bc9b4cca7caa5a4acc3c936471da2ddd258f74bd4c03cf64a7d1c975954f95"},"schema_version":"1.0"},"canonical_sha256":"5f41719b6fa7aae2901024191461d8f1460affcaef6722aa3a39064d71843e76","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:53:10.623860Z","signature_b64":"KDUuwzTma/ttyoUr77ZJjvDwDFXDU83oFQgHEOFrPLxfsuE5hxwvcMdJzjJAEj8x9CvPb1QvMccE2jpIaFZTAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5f41719b6fa7aae2901024191461d8f1460affcaef6722aa3a39064d71843e76","last_reissued_at":"2026-07-05T10:53:10.623392Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:53:10.623392Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.14088","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:53:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4KI1Gt8ApyXvGYXo9Pax/YKvJ1mPKb6ZzBxWv57qavUl5L5EP1mzNQ2sg4BhneV+p89iBjfKpFYB+mjC/G7FBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-14T09:59:58.084473Z"},"content_sha256":"442610a029e24a923dff866dbc99760aa21fd62da1a7857cae7f4316d90f7abc","schema_version":"1.0","event_id":"sha256:442610a029e24a923dff866dbc99760aa21fd62da1a7857cae7f4316d90f7abc"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:L5AXDG3PU6VOFEAQEQMRIYOY6F","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ReaL: Efficient RLHF Training of Large Language Models with Parameter Reallocation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.DC","authors_text":"Guangju Wang, Huanchen Zhang, Kaiwei Li, Wei Fu, Yi Wu, Zhiyu Mei","submitted_at":"2024-06-20T08:04:07Z","abstract_excerpt":"Reinforcement Learning from Human Feedback (RLHF) is a pivotal technique for empowering large language model (LLM) applications. Compared with the supervised training process of LLMs, the RLHF training process is much more sophisticated, requiring a diverse range of computation workloads with intricate dependencies between multiple LLM instances. Therefore, simply adopting the fixed parallelization strategies from supervised training for LLMs can be insufficient for RLHF and result in low training efficiency. To overcome this limitation, we propose a novel technique named parameter ReaLlocatio"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.14088","kind":"arxiv","version":2},"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/2406.14088/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":"2026-07-05T10:53:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"e8bPPEx8rxkkJwFkNBKr48GGoMoqim4pIpvmUaSwydcS35XQvIpj0EYeN+dBg52RtruHosPS+JOheKK1D1yECA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-14T09:59:58.084855Z"},"content_sha256":"db10140ebcf003b2df8f5551eb5cb09b45fac5eccc4f3a2a8e83486f97b83d6b","schema_version":"1.0","event_id":"sha256:db10140ebcf003b2df8f5551eb5cb09b45fac5eccc4f3a2a8e83486f97b83d6b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/L5AXDG3PU6VOFEAQEQMRIYOY6F/bundle.json","state_url":"https://pith.science/pith/L5AXDG3PU6VOFEAQEQMRIYOY6F/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/L5AXDG3PU6VOFEAQEQMRIYOY6F/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":"2026-07-14T09:59:58Z","links":{"resolver":"https://pith.science/pith/L5AXDG3PU6VOFEAQEQMRIYOY6F","bundle":"https://pith.science/pith/L5AXDG3PU6VOFEAQEQMRIYOY6F/bundle.json","state":"https://pith.science/pith/L5AXDG3PU6VOFEAQEQMRIYOY6F/state.json","well_known_bundle":"https://pith.science/.well-known/pith/L5AXDG3PU6VOFEAQEQMRIYOY6F/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:L5AXDG3PU6VOFEAQEQMRIYOY6F","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":"11bc9b4cca7caa5a4acc3c936471da2ddd258f74bd4c03cf64a7d1c975954f95","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2024-06-20T08:04:07Z","title_canon_sha256":"da7ec43cbeb19100f4ef4804d02de73984a8aa2b0ed119a868b59912cfe19823"},"schema_version":"1.0","source":{"id":"2406.14088","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.14088","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"arxiv_version","alias_value":"2406.14088v2","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.14088","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"pith_short_12","alias_value":"L5AXDG3PU6VO","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"pith_short_16","alias_value":"L5AXDG3PU6VOFEAQ","created_at":"2026-07-05T10:53:10Z"},{"alias_kind":"pith_short_8","alias_value":"L5AXDG3P","created_at":"2026-07-05T10:53:10Z"}],"graph_snapshots":[{"event_id":"sha256:db10140ebcf003b2df8f5551eb5cb09b45fac5eccc4f3a2a8e83486f97b83d6b","target":"graph","created_at":"2026-07-05T10:53:10Z","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/2406.14088/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement Learning from Human Feedback (RLHF) is a pivotal technique for empowering large language model (LLM) applications. Compared with the supervised training process of LLMs, the RLHF training process is much more sophisticated, requiring a diverse range of computation workloads with intricate dependencies between multiple LLM instances. Therefore, simply adopting the fixed parallelization strategies from supervised training for LLMs can be insufficient for RLHF and result in low training efficiency. To overcome this limitation, we propose a novel technique named parameter ReaLlocatio","authors_text":"Guangju Wang, Huanchen Zhang, Kaiwei Li, Wei Fu, Yi Wu, Zhiyu Mei","cross_cats":["cs.AI","cs.CL","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2024-06-20T08:04:07Z","title":"ReaL: Efficient RLHF Training of Large Language Models with Parameter Reallocation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.14088","kind":"arxiv","version":2},"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":"sha256:442610a029e24a923dff866dbc99760aa21fd62da1a7857cae7f4316d90f7abc","target":"record","created_at":"2026-07-05T10:53:10Z","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":"11bc9b4cca7caa5a4acc3c936471da2ddd258f74bd4c03cf64a7d1c975954f95","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2024-06-20T08:04:07Z","title_canon_sha256":"da7ec43cbeb19100f4ef4804d02de73984a8aa2b0ed119a868b59912cfe19823"},"schema_version":"1.0","source":{"id":"2406.14088","kind":"arxiv","version":2}},"canonical_sha256":"5f41719b6fa7aae2901024191461d8f1460affcaef6722aa3a39064d71843e76","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5f41719b6fa7aae2901024191461d8f1460affcaef6722aa3a39064d71843e76","first_computed_at":"2026-07-05T10:53:10.623392Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:53:10.623392Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"KDUuwzTma/ttyoUr77ZJjvDwDFXDU83oFQgHEOFrPLxfsuE5hxwvcMdJzjJAEj8x9CvPb1QvMccE2jpIaFZTAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:53:10.623860Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.14088","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:442610a029e24a923dff866dbc99760aa21fd62da1a7857cae7f4316d90f7abc","sha256:db10140ebcf003b2df8f5551eb5cb09b45fac5eccc4f3a2a8e83486f97b83d6b"],"state_sha256":"c6512a6ad8e8f9568eaea00e8378139766d17b79b62896823396806321125a6e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8WXrRxo0PU+tI5xJtXw1zOxsiEwonsdAJIqc3yvoLgEcS8qSEpovZ0EejN4+7LM9C1UbuXk30fpKkLzPPAa0Aw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-14T09:59:58.087475Z","bundle_sha256":"94647a55bb0256c0be14aed851ca3e058f77287bda88b0f494b998b65926d55e"}}