{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:RJO5WKAI3KS4SWUY33DE5KRF5B","short_pith_number":"pith:RJO5WKAI","canonical_record":{"source":{"id":"2406.01495","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T16:21:38Z","cross_cats_sorted":[],"title_canon_sha256":"4a8bf708b9d7a0809a45a27172eb17dc10e02092d7501d10d12d5c6e8345ed9c","abstract_canon_sha256":"1cff10719c2a142263036c39828150daf21047fa4de4dafcff66b660f612c0ac"},"schema_version":"1.0"},"canonical_sha256":"8a5ddb2808daa5c95a98dec64eaa25e8484291ab91ce435fd924d4a8d42f74e5","source":{"kind":"arxiv","id":"2406.01495","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.01495","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"arxiv_version","alias_value":"2406.01495v3","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.01495","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"pith_short_12","alias_value":"RJO5WKAI3KS4","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"pith_short_16","alias_value":"RJO5WKAI3KS4SWUY","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"pith_short_8","alias_value":"RJO5WKAI","created_at":"2026-07-05T10:59:23Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:RJO5WKAI3KS4SWUY33DE5KRF5B","target":"record","payload":{"canonical_record":{"source":{"id":"2406.01495","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T16:21:38Z","cross_cats_sorted":[],"title_canon_sha256":"4a8bf708b9d7a0809a45a27172eb17dc10e02092d7501d10d12d5c6e8345ed9c","abstract_canon_sha256":"1cff10719c2a142263036c39828150daf21047fa4de4dafcff66b660f612c0ac"},"schema_version":"1.0"},"canonical_sha256":"8a5ddb2808daa5c95a98dec64eaa25e8484291ab91ce435fd924d4a8d42f74e5","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:59:23.774376Z","signature_b64":"oL9ceh0FVkwu+yOFl4w/frPQewfrIiIgTzsJQ6OC/9wk+gHMsThdUjIFc6DRBt0qQNermSSoXSPIZJCYjHRdDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8a5ddb2808daa5c95a98dec64eaa25e8484291ab91ce435fd924d4a8d42f74e5","last_reissued_at":"2026-07-05T10:59:23.773858Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:59:23.773858Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2406.01495","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:59:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"P2vJIRU+YjDX07alhJsKPTNlOb+6oYLzQG8PdZCA/NTnkmoNv9ozczRMo2arauG7UxkAMKvQBN0iHZt2upWECQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T17:24:48.682267Z"},"content_sha256":"bd37be3233e09e36c3e1604e15271046b763505c0184725d2912d977d6d52ee1","schema_version":"1.0","event_id":"sha256:bd37be3233e09e36c3e1604e15271046b763505c0184725d2912d977d6d52ee1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:RJO5WKAI3KS4SWUY33DE5KRF5B","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Re-ReST: Reflection-Reinforced Self-Training for Language Agents","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Cheng-Fu Yang, Kai-Wei Chang, Nanyun Peng, Xueqing Wu, Zi-Yi Dou","submitted_at":"2024-06-03T16:21:38Z","abstract_excerpt":"Finetuning language agents with reasoning-action trajectories is effective, but obtaining these trajectories from human annotations or stronger models is costly and sometimes impractical. In this paper, we investigate the use of self-training in language agents, which can generate supervision from the agent itself, offering a promising alternative without relying on human or stronger model demonstrations. Self-training, however, requires high-quality model-generated samples, which are hard to obtain for challenging language agent tasks. To address this, we present Reflection-Reinforced Self-Tr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.01495","kind":"arxiv","version":3},"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/2406.01495/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":"2026-07-05T10:59:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Pcp4KxuuSTWBdkqc82HJKN+ivJv/Qd2wN8J5RwK4SJ5oAO50TIeyaQu6y03YLQCf/VLR+GVhwZRUsYWyaiI0BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T17:24:48.682645Z"},"content_sha256":"f3aa209e8330e7cb14a271b3cc9d807117abf2a2489c6bca81d6cde578f8f406","schema_version":"1.0","event_id":"sha256:f3aa209e8330e7cb14a271b3cc9d807117abf2a2489c6bca81d6cde578f8f406"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/RJO5WKAI3KS4SWUY33DE5KRF5B/bundle.json","state_url":"https://pith.science/pith/RJO5WKAI3KS4SWUY33DE5KRF5B/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/RJO5WKAI3KS4SWUY33DE5KRF5B/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":"2026-07-06T17:24:48Z","links":{"resolver":"https://pith.science/pith/RJO5WKAI3KS4SWUY33DE5KRF5B","bundle":"https://pith.science/pith/RJO5WKAI3KS4SWUY33DE5KRF5B/bundle.json","state":"https://pith.science/pith/RJO5WKAI3KS4SWUY33DE5KRF5B/state.json","well_known_bundle":"https://pith.science/.well-known/pith/RJO5WKAI3KS4SWUY33DE5KRF5B/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:RJO5WKAI3KS4SWUY33DE5KRF5B","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":"1cff10719c2a142263036c39828150daf21047fa4de4dafcff66b660f612c0ac","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T16:21:38Z","title_canon_sha256":"4a8bf708b9d7a0809a45a27172eb17dc10e02092d7501d10d12d5c6e8345ed9c"},"schema_version":"1.0","source":{"id":"2406.01495","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.01495","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"arxiv_version","alias_value":"2406.01495v3","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.01495","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"pith_short_12","alias_value":"RJO5WKAI3KS4","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"pith_short_16","alias_value":"RJO5WKAI3KS4SWUY","created_at":"2026-07-05T10:59:23Z"},{"alias_kind":"pith_short_8","alias_value":"RJO5WKAI","created_at":"2026-07-05T10:59:23Z"}],"graph_snapshots":[{"event_id":"sha256:f3aa209e8330e7cb14a271b3cc9d807117abf2a2489c6bca81d6cde578f8f406","target":"graph","created_at":"2026-07-05T10:59:23Z","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/2406.01495/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Finetuning language agents with reasoning-action trajectories is effective, but obtaining these trajectories from human annotations or stronger models is costly and sometimes impractical. In this paper, we investigate the use of self-training in language agents, which can generate supervision from the agent itself, offering a promising alternative without relying on human or stronger model demonstrations. Self-training, however, requires high-quality model-generated samples, which are hard to obtain for challenging language agent tasks. To address this, we present Reflection-Reinforced Self-Tr","authors_text":"Cheng-Fu Yang, Kai-Wei Chang, Nanyun Peng, Xueqing Wu, Zi-Yi Dou","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T16:21:38Z","title":"Re-ReST: Reflection-Reinforced Self-Training for Language Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.01495","kind":"arxiv","version":3},"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:bd37be3233e09e36c3e1604e15271046b763505c0184725d2912d977d6d52ee1","target":"record","created_at":"2026-07-05T10:59:23Z","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":"1cff10719c2a142263036c39828150daf21047fa4de4dafcff66b660f612c0ac","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-03T16:21:38Z","title_canon_sha256":"4a8bf708b9d7a0809a45a27172eb17dc10e02092d7501d10d12d5c6e8345ed9c"},"schema_version":"1.0","source":{"id":"2406.01495","kind":"arxiv","version":3}},"canonical_sha256":"8a5ddb2808daa5c95a98dec64eaa25e8484291ab91ce435fd924d4a8d42f74e5","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8a5ddb2808daa5c95a98dec64eaa25e8484291ab91ce435fd924d4a8d42f74e5","first_computed_at":"2026-07-05T10:59:23.773858Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:59:23.773858Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"oL9ceh0FVkwu+yOFl4w/frPQewfrIiIgTzsJQ6OC/9wk+gHMsThdUjIFc6DRBt0qQNermSSoXSPIZJCYjHRdDg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:59:23.774376Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.01495","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:bd37be3233e09e36c3e1604e15271046b763505c0184725d2912d977d6d52ee1","sha256:f3aa209e8330e7cb14a271b3cc9d807117abf2a2489c6bca81d6cde578f8f406"],"state_sha256":"82c9605fcb2e220a990ac632e5112c62aa3498d9c715dec537c2ea6afbb29118"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XCPxyb4MPGeGM+g7MmHrmmICi92pnE/LoZqNyCFf0/C2poXFaLkpsuAUBHCg6zTEI/xrWDhuFrwf5+wzkjx9Ag==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T17:24:48.684684Z","bundle_sha256":"01f127c0cceceab8469009b0367bf836a3b6f0d7f2e183f584d0ac62b6d9dbb9"}}