{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:VWFEEQU7WZ4H3NUNUBXJARMD4R","short_pith_number":"pith:VWFEEQU7","canonical_record":{"source":{"id":"2608.03972","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-04T17:40:08Z","cross_cats_sorted":[],"title_canon_sha256":"baa1c1c5459084ca240ff90c6763a8890f2569f4629f98dbcfb720a82d0853b1","abstract_canon_sha256":"558863715f3ef64e0df1b252d675268bb32a5df00b2d36900eea1b04d183ac9e"},"schema_version":"1.0"},"canonical_sha256":"ad8a42429fb6787db68da06e904583e46b5c3349d01b93f22d8ecc8daeb8d723","source":{"kind":"arxiv","id":"2608.03972","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.03972","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"arxiv_version","alias_value":"2608.03972v1","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.03972","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"pith_short_12","alias_value":"VWFEEQU7WZ4H","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"pith_short_16","alias_value":"VWFEEQU7WZ4H3NUN","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"pith_short_8","alias_value":"VWFEEQU7","created_at":"2026-08-05T01:37:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:VWFEEQU7WZ4H3NUNUBXJARMD4R","target":"record","payload":{"canonical_record":{"source":{"id":"2608.03972","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-04T17:40:08Z","cross_cats_sorted":[],"title_canon_sha256":"baa1c1c5459084ca240ff90c6763a8890f2569f4629f98dbcfb720a82d0853b1","abstract_canon_sha256":"558863715f3ef64e0df1b252d675268bb32a5df00b2d36900eea1b04d183ac9e"},"schema_version":"1.0"},"canonical_sha256":"ad8a42429fb6787db68da06e904583e46b5c3349d01b93f22d8ecc8daeb8d723","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-05T01:37:37.038904Z","signature_b64":"grWXkvaYkxe7Ce/U2prCAdssLVrMCrKBuLiGb7KPcKUg5lPDD/l3RW3MPisd/qI7u0D0qEMx2GprmQ7gHoYsBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ad8a42429fb6787db68da06e904583e46b5c3349d01b93f22d8ecc8daeb8d723","last_reissued_at":"2026-08-05T01:37:37.037377Z","signature_status":"signed_v1","first_computed_at":"2026-08-05T01:37:37.037377Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.03972","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":"2026-08-05T01:37:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"03yUSgEgRXhPpZ79vOyku2ar3ta9yq4SKNLNf/pzrucOw8Kw9lLfJi/EZ0vFxen5f/UaJjfU++e8aP0pmRrgDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T12:59:32.027316Z"},"content_sha256":"5d692c89b1e634840ab907491d800f24079711648719b94f58d176395c9f53f6","schema_version":"1.0","event_id":"sha256:5d692c89b1e634840ab907491d800f24079711648719b94f58d176395c9f53f6"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:VWFEEQU7WZ4H3NUNUBXJARMD4R","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ReflectRL: Learning from Golden Negative Trajectories via Reflective-to-Direct Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Aniri, Chennan Zhou, Fei Shen, Hu Cao, Jinhe Bi, Shuo Lu, Tat-Seng Chua, Volker Tresp, Wenke Huang, Xun Xiao, Yunpu Ma, Zengjie Jin, Zhihong Zhu","submitted_at":"2026-08-04T17:40:08Z","abstract_excerpt":"On-policy training has emerged as a powerful post-training paradigm for improving the reasoning capabilities of large language models, and is often enhanced by golden trajectories from stronger expert models. However, when the expert fails on harder problems, existing trajectory-guided methods lose their main source of supervision, and these failed trajectories are typically discarded as negative samples. We argue that such failures, which we call Golden Negative Trajectories, can still provide valuable reasoning signals when treated not as demonstrations to imitate, but as flawed trajectories"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.03972","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/2608.03972/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":"2026-08-05T01:37:37Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2Z/7Mb9RDqEostxnmBqkmhm9ujGfMLvuk175sW0yovjTXOoGh2DtqWVsxMn28h3FawvhnuKNolJlgfyCyQH7AA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T12:59:32.027833Z"},"content_sha256":"d7060a44df816d14da99f23fb59e66877f91a6770e659e4534a6db5b723e7d3f","schema_version":"1.0","event_id":"sha256:d7060a44df816d14da99f23fb59e66877f91a6770e659e4534a6db5b723e7d3f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VWFEEQU7WZ4H3NUNUBXJARMD4R/bundle.json","state_url":"https://pith.science/pith/VWFEEQU7WZ4H3NUNUBXJARMD4R/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VWFEEQU7WZ4H3NUNUBXJARMD4R/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":"2026-08-05T12:59:32Z","links":{"resolver":"https://pith.science/pith/VWFEEQU7WZ4H3NUNUBXJARMD4R","bundle":"https://pith.science/pith/VWFEEQU7WZ4H3NUNUBXJARMD4R/bundle.json","state":"https://pith.science/pith/VWFEEQU7WZ4H3NUNUBXJARMD4R/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VWFEEQU7WZ4H3NUNUBXJARMD4R/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:VWFEEQU7WZ4H3NUNUBXJARMD4R","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":"558863715f3ef64e0df1b252d675268bb32a5df00b2d36900eea1b04d183ac9e","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-04T17:40:08Z","title_canon_sha256":"baa1c1c5459084ca240ff90c6763a8890f2569f4629f98dbcfb720a82d0853b1"},"schema_version":"1.0","source":{"id":"2608.03972","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.03972","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"arxiv_version","alias_value":"2608.03972v1","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.03972","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"pith_short_12","alias_value":"VWFEEQU7WZ4H","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"pith_short_16","alias_value":"VWFEEQU7WZ4H3NUN","created_at":"2026-08-05T01:37:37Z"},{"alias_kind":"pith_short_8","alias_value":"VWFEEQU7","created_at":"2026-08-05T01:37:37Z"}],"graph_snapshots":[{"event_id":"sha256:d7060a44df816d14da99f23fb59e66877f91a6770e659e4534a6db5b723e7d3f","target":"graph","created_at":"2026-08-05T01:37:37Z","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/2608.03972/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"On-policy training has emerged as a powerful post-training paradigm for improving the reasoning capabilities of large language models, and is often enhanced by golden trajectories from stronger expert models. However, when the expert fails on harder problems, existing trajectory-guided methods lose their main source of supervision, and these failed trajectories are typically discarded as negative samples. We argue that such failures, which we call Golden Negative Trajectories, can still provide valuable reasoning signals when treated not as demonstrations to imitate, but as flawed trajectories","authors_text":"Aniri, Chennan Zhou, Fei Shen, Hu Cao, Jinhe Bi, Shuo Lu, Tat-Seng Chua, Volker Tresp, Wenke Huang, Xun Xiao, Yunpu Ma, Zengjie Jin, Zhihong Zhu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-04T17:40:08Z","title":"ReflectRL: Learning from Golden Negative Trajectories via Reflective-to-Direct Reasoning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.03972","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":"sha256:5d692c89b1e634840ab907491d800f24079711648719b94f58d176395c9f53f6","target":"record","created_at":"2026-08-05T01:37:37Z","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":"558863715f3ef64e0df1b252d675268bb32a5df00b2d36900eea1b04d183ac9e","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-08-04T17:40:08Z","title_canon_sha256":"baa1c1c5459084ca240ff90c6763a8890f2569f4629f98dbcfb720a82d0853b1"},"schema_version":"1.0","source":{"id":"2608.03972","kind":"arxiv","version":1}},"canonical_sha256":"ad8a42429fb6787db68da06e904583e46b5c3349d01b93f22d8ecc8daeb8d723","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ad8a42429fb6787db68da06e904583e46b5c3349d01b93f22d8ecc8daeb8d723","first_computed_at":"2026-08-05T01:37:37.037377Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-05T01:37:37.037377Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"grWXkvaYkxe7Ce/U2prCAdssLVrMCrKBuLiGb7KPcKUg5lPDD/l3RW3MPisd/qI7u0D0qEMx2GprmQ7gHoYsBg==","signature_status":"signed_v1","signed_at":"2026-08-05T01:37:37.038904Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.03972","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5d692c89b1e634840ab907491d800f24079711648719b94f58d176395c9f53f6","sha256:d7060a44df816d14da99f23fb59e66877f91a6770e659e4534a6db5b723e7d3f"],"state_sha256":"e8d148ec7ba66659baddc96160f66590897605d225895d344fc297e8ca356309"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dwOvVdjlQue78uUqeDwHpAdrOFEBhYGK8QJdXf7mx3qp1I/w2jIxVXZyBYB4nTMy0COT5RZ2EIcw2Ps41BLPAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T12:59:32.033056Z","bundle_sha256":"aabbd5df9f9a7f9b3ad723345a25e86a338e58870a3e3c8c0339248d80711a21"}}