{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:3V2WUZTWJ3UY6VTNHWZKTIJO2P","short_pith_number":"pith:3V2WUZTW","canonical_record":{"source":{"id":"2608.05541","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T02:39:10Z","cross_cats_sorted":[],"title_canon_sha256":"d83f4aac43d2bacdf99413d814c60dccaca879be6eeab59e497e9cde088fa204","abstract_canon_sha256":"31ede34a048041441fb2d0d3aa1f3a81e1439b5e59df065cd1a7c3e2661d593f"},"schema_version":"1.0"},"canonical_sha256":"dd756a66764ee98f566d3db2a9a12ed3eeeff2f56d3306127ccabdc23b8267bb","source":{"kind":"arxiv","id":"2608.05541","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.05541","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"arxiv_version","alias_value":"2608.05541v1","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.05541","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"pith_short_12","alias_value":"3V2WUZTWJ3UY","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"pith_short_16","alias_value":"3V2WUZTWJ3UY6VTN","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"pith_short_8","alias_value":"3V2WUZTW","created_at":"2026-08-07T00:48:36Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:3V2WUZTWJ3UY6VTNHWZKTIJO2P","target":"record","payload":{"canonical_record":{"source":{"id":"2608.05541","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T02:39:10Z","cross_cats_sorted":[],"title_canon_sha256":"d83f4aac43d2bacdf99413d814c60dccaca879be6eeab59e497e9cde088fa204","abstract_canon_sha256":"31ede34a048041441fb2d0d3aa1f3a81e1439b5e59df065cd1a7c3e2661d593f"},"schema_version":"1.0"},"canonical_sha256":"dd756a66764ee98f566d3db2a9a12ed3eeeff2f56d3306127ccabdc23b8267bb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-07T00:48:36.635752Z","signature_b64":"vdmViaUeHcqOIuNOpDh3x0s6yakofxt4G45CByuZc9RdYPoDzxeIkiT85HO2lfNoT1N9Fg1nhNpjYhjEUVNUBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dd756a66764ee98f566d3db2a9a12ed3eeeff2f56d3306127ccabdc23b8267bb","last_reissued_at":"2026-08-07T00:48:36.634323Z","signature_status":"signed_v1","first_computed_at":"2026-08-07T00:48:36.634323Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.05541","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":"2026-08-07T00:48:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JDUApQHwbLC/RVr+LkaG7/CMgKUEqVQT+pn/i9vxhA2IcpQP2LdgIfQ2iFwoRZYH8pcIT3aZt0MkK9tw5nuiCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T18:13:13.229994Z"},"content_sha256":"5fb7b8776f0e43eaef514523ae82c63ac8721d122459e3bf68ade5d32c564c77","schema_version":"1.0","event_id":"sha256:5fb7b8776f0e43eaef514523ae82c63ac8721d122459e3bf68ade5d32c564c77"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:3V2WUZTWJ3UY6VTNHWZKTIJO2P","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Hyper-ES: Effective Evolution Strategies for LLM Reasoning via Descent Direction Merging","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Mingxuan Yuan, Xialiang Tong, Yu Gu, Yunpeng Ba, Zhenkun Wang, Zhi Zheng","submitted_at":"2026-08-06T02:39:10Z","abstract_excerpt":"Evolution Strategy (ES) is a promising alternative to gradient-based fine-tuning for resource-constrained Large Language Model (LLM) reasoning. However, directly applying ES to billion-parameter LLMs is highly ineffective. In such high-dimensional parameter spaces, most random perturbations are nearly orthogonal to useful update directions, leading to unstable optimization. We propose Hyper-ES, a subspace-based ES framework that avoids the weakness of ES in full-parameter search while exploiting its strength in low-dimensional optimization. Instead of asking ES to discover useful directions fr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.05541","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/2608.05541/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":"2026-08-07T00:48:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"H1oVz+jNY2iBi2TXSmmoH0QnghQRmAaV68yakicCmF/03fisFeLYXJioKmceH/9Bymm2S3PucG8yZO9VdlkfAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T18:13:13.230942Z"},"content_sha256":"d0758391e48c9b0d27d07cc3a6464cdde90957b5ec5135f26d6a884b3c439c9d","schema_version":"1.0","event_id":"sha256:d0758391e48c9b0d27d07cc3a6464cdde90957b5ec5135f26d6a884b3c439c9d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3V2WUZTWJ3UY6VTNHWZKTIJO2P/bundle.json","state_url":"https://pith.science/pith/3V2WUZTWJ3UY6VTNHWZKTIJO2P/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3V2WUZTWJ3UY6VTNHWZKTIJO2P/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":"2026-08-08T18:13:13Z","links":{"resolver":"https://pith.science/pith/3V2WUZTWJ3UY6VTNHWZKTIJO2P","bundle":"https://pith.science/pith/3V2WUZTWJ3UY6VTNHWZKTIJO2P/bundle.json","state":"https://pith.science/pith/3V2WUZTWJ3UY6VTNHWZKTIJO2P/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3V2WUZTWJ3UY6VTNHWZKTIJO2P/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:3V2WUZTWJ3UY6VTNHWZKTIJO2P","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":"31ede34a048041441fb2d0d3aa1f3a81e1439b5e59df065cd1a7c3e2661d593f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T02:39:10Z","title_canon_sha256":"d83f4aac43d2bacdf99413d814c60dccaca879be6eeab59e497e9cde088fa204"},"schema_version":"1.0","source":{"id":"2608.05541","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.05541","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"arxiv_version","alias_value":"2608.05541v1","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.05541","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"pith_short_12","alias_value":"3V2WUZTWJ3UY","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"pith_short_16","alias_value":"3V2WUZTWJ3UY6VTN","created_at":"2026-08-07T00:48:36Z"},{"alias_kind":"pith_short_8","alias_value":"3V2WUZTW","created_at":"2026-08-07T00:48:36Z"}],"graph_snapshots":[{"event_id":"sha256:d0758391e48c9b0d27d07cc3a6464cdde90957b5ec5135f26d6a884b3c439c9d","target":"graph","created_at":"2026-08-07T00:48:36Z","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/2608.05541/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Evolution Strategy (ES) is a promising alternative to gradient-based fine-tuning for resource-constrained Large Language Model (LLM) reasoning. However, directly applying ES to billion-parameter LLMs is highly ineffective. In such high-dimensional parameter spaces, most random perturbations are nearly orthogonal to useful update directions, leading to unstable optimization. We propose Hyper-ES, a subspace-based ES framework that avoids the weakness of ES in full-parameter search while exploiting its strength in low-dimensional optimization. Instead of asking ES to discover useful directions fr","authors_text":"Mingxuan Yuan, Xialiang Tong, Yu Gu, Yunpeng Ba, Zhenkun Wang, Zhi Zheng","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T02:39:10Z","title":"Hyper-ES: Effective Evolution Strategies for LLM Reasoning via Descent Direction Merging"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.05541","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":"sha256:5fb7b8776f0e43eaef514523ae82c63ac8721d122459e3bf68ade5d32c564c77","target":"record","created_at":"2026-08-07T00:48:36Z","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":"31ede34a048041441fb2d0d3aa1f3a81e1439b5e59df065cd1a7c3e2661d593f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-08-06T02:39:10Z","title_canon_sha256":"d83f4aac43d2bacdf99413d814c60dccaca879be6eeab59e497e9cde088fa204"},"schema_version":"1.0","source":{"id":"2608.05541","kind":"arxiv","version":1}},"canonical_sha256":"dd756a66764ee98f566d3db2a9a12ed3eeeff2f56d3306127ccabdc23b8267bb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dd756a66764ee98f566d3db2a9a12ed3eeeff2f56d3306127ccabdc23b8267bb","first_computed_at":"2026-08-07T00:48:36.634323Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-07T00:48:36.634323Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vdmViaUeHcqOIuNOpDh3x0s6yakofxt4G45CByuZc9RdYPoDzxeIkiT85HO2lfNoT1N9Fg1nhNpjYhjEUVNUBg==","signature_status":"signed_v1","signed_at":"2026-08-07T00:48:36.635752Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.05541","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5fb7b8776f0e43eaef514523ae82c63ac8721d122459e3bf68ade5d32c564c77","sha256:d0758391e48c9b0d27d07cc3a6464cdde90957b5ec5135f26d6a884b3c439c9d"],"state_sha256":"aba395abc53e95c1ae28aa8f75f5a0373770d960612b2c88338f6090c115d47b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GY6jOxVetltHzXdTQg/uFKe5lkWIoWmfa2GsIwVeIunidJ9zFGmkhZGSUb1kLyC05rd/ky7Rcf+IJBcmoYEXDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T18:13:13.244211Z","bundle_sha256":"28919206617aef50c6933c99019bc89f2045b305bb740a0ac45e1d348febd2d6"}}