{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:G2PFPF7HQD53RIAKIV6SN7ICC4","short_pith_number":"pith:G2PFPF7H","canonical_record":{"source":{"id":"2206.10498","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-06-21T16:15:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1608a2d221e1d73b9e1f4e7a2f921e7350942960c030beb6e397792a67c56a74","abstract_canon_sha256":"d9ca01acd7f6015f0bd7f8fd15a0867f25fc6e37a642e885e88c129472acbe3c"},"schema_version":"1.0"},"canonical_sha256":"369e5797e780fbb8a00a457d26fd0217114f82b4d579e648d13e95f33c3719bb","source":{"kind":"arxiv","id":"2206.10498","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2206.10498","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"arxiv_version","alias_value":"2206.10498v4","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.10498","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"pith_short_12","alias_value":"G2PFPF7HQD53","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"pith_short_16","alias_value":"G2PFPF7HQD53RIAK","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"pith_short_8","alias_value":"G2PFPF7H","created_at":"2026-07-05T07:16:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:G2PFPF7HQD53RIAKIV6SN7ICC4","target":"record","payload":{"canonical_record":{"source":{"id":"2206.10498","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-06-21T16:15:27Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"1608a2d221e1d73b9e1f4e7a2f921e7350942960c030beb6e397792a67c56a74","abstract_canon_sha256":"d9ca01acd7f6015f0bd7f8fd15a0867f25fc6e37a642e885e88c129472acbe3c"},"schema_version":"1.0"},"canonical_sha256":"369e5797e780fbb8a00a457d26fd0217114f82b4d579e648d13e95f33c3719bb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:16:32.299337Z","signature_b64":"iDXP48g07pOOW5hoxEbK2bHtLhYpCkhfB2vmdbrC+i6nYehwS6HLSi+UnWjhL1wSyqD6fo2Cj4qfHx77rGzCCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"369e5797e780fbb8a00a457d26fd0217114f82b4d579e648d13e95f33c3719bb","last_reissued_at":"2026-07-05T07:16:32.298785Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:16:32.298785Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2206.10498","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T07:16:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5Y3MUenxqEblZ5WzVqkU9kSKy43Xbug7JSyNAPevJMVRjFMVoOkTcnudr/oLlcls/YgDc1BW9Z8yPvGsnorrBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T18:52:16.575311Z"},"content_sha256":"dd311495da86dad5736aaced04ae5ebdfd239cac5be677e40bf6e9ee3a5f70d8","schema_version":"1.0","event_id":"sha256:dd311495da86dad5736aaced04ae5ebdfd239cac5be677e40bf6e9ee3a5f70d8"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:G2PFPF7HQD53RIAKIV6SN7ICC4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"PlanBench: An Extensible Benchmark for Evaluating Large Language Models on Planning and Reasoning about Change","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Alberto Olmo, Karthik Valmeekam, Matthew Marquez, Sarath Sreedharan, Subbarao Kambhampati","submitted_at":"2022-06-21T16:15:27Z","abstract_excerpt":"Generating plans of action, and reasoning about change have long been considered a core competence of intelligent agents. It is thus no surprise that evaluating the planning and reasoning capabilities of large language models (LLMs) has become a hot topic of research. Most claims about LLM planning capabilities are however based on common sense tasks-where it becomes hard to tell whether LLMs are planning or merely retrieving from their vast world knowledge. There is a strong need for systematic and extensible planning benchmarks with sufficient diversity to evaluate whether LLMs have innate p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.10498","kind":"arxiv","version":4},"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/2206.10498/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":"2026-07-05T07:16:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"0Fxs8SzlHJd/qjzXY6ce2TOZV/RyiwoAb5iYI7P9NC4tYsAPHMR3dGA9eSwM8ieBnxaYOpudAv7ZY36uzsa2CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T18:52:16.575848Z"},"content_sha256":"3617cc18d4b30a7eaa65387b5efcc55a5f979cf37528d506519d2eabc634f98b","schema_version":"1.0","event_id":"sha256:3617cc18d4b30a7eaa65387b5efcc55a5f979cf37528d506519d2eabc634f98b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/G2PFPF7HQD53RIAKIV6SN7ICC4/bundle.json","state_url":"https://pith.science/pith/G2PFPF7HQD53RIAKIV6SN7ICC4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/G2PFPF7HQD53RIAKIV6SN7ICC4/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":"2026-08-03T18:52:16Z","links":{"resolver":"https://pith.science/pith/G2PFPF7HQD53RIAKIV6SN7ICC4","bundle":"https://pith.science/pith/G2PFPF7HQD53RIAKIV6SN7ICC4/bundle.json","state":"https://pith.science/pith/G2PFPF7HQD53RIAKIV6SN7ICC4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/G2PFPF7HQD53RIAKIV6SN7ICC4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:G2PFPF7HQD53RIAKIV6SN7ICC4","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":"d9ca01acd7f6015f0bd7f8fd15a0867f25fc6e37a642e885e88c129472acbe3c","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-06-21T16:15:27Z","title_canon_sha256":"1608a2d221e1d73b9e1f4e7a2f921e7350942960c030beb6e397792a67c56a74"},"schema_version":"1.0","source":{"id":"2206.10498","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2206.10498","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"arxiv_version","alias_value":"2206.10498v4","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2206.10498","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"pith_short_12","alias_value":"G2PFPF7HQD53","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"pith_short_16","alias_value":"G2PFPF7HQD53RIAK","created_at":"2026-07-05T07:16:32Z"},{"alias_kind":"pith_short_8","alias_value":"G2PFPF7H","created_at":"2026-07-05T07:16:32Z"}],"graph_snapshots":[{"event_id":"sha256:3617cc18d4b30a7eaa65387b5efcc55a5f979cf37528d506519d2eabc634f98b","target":"graph","created_at":"2026-07-05T07:16:32Z","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/2206.10498/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Generating plans of action, and reasoning about change have long been considered a core competence of intelligent agents. It is thus no surprise that evaluating the planning and reasoning capabilities of large language models (LLMs) has become a hot topic of research. Most claims about LLM planning capabilities are however based on common sense tasks-where it becomes hard to tell whether LLMs are planning or merely retrieving from their vast world knowledge. There is a strong need for systematic and extensible planning benchmarks with sufficient diversity to evaluate whether LLMs have innate p","authors_text":"Alberto Olmo, Karthik Valmeekam, Matthew Marquez, Sarath Sreedharan, Subbarao Kambhampati","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-06-21T16:15:27Z","title":"PlanBench: An Extensible Benchmark for Evaluating Large Language Models on Planning and Reasoning about Change"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2206.10498","kind":"arxiv","version":4},"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:dd311495da86dad5736aaced04ae5ebdfd239cac5be677e40bf6e9ee3a5f70d8","target":"record","created_at":"2026-07-05T07:16:32Z","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":"d9ca01acd7f6015f0bd7f8fd15a0867f25fc6e37a642e885e88c129472acbe3c","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-06-21T16:15:27Z","title_canon_sha256":"1608a2d221e1d73b9e1f4e7a2f921e7350942960c030beb6e397792a67c56a74"},"schema_version":"1.0","source":{"id":"2206.10498","kind":"arxiv","version":4}},"canonical_sha256":"369e5797e780fbb8a00a457d26fd0217114f82b4d579e648d13e95f33c3719bb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"369e5797e780fbb8a00a457d26fd0217114f82b4d579e648d13e95f33c3719bb","first_computed_at":"2026-07-05T07:16:32.298785Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:16:32.298785Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"iDXP48g07pOOW5hoxEbK2bHtLhYpCkhfB2vmdbrC+i6nYehwS6HLSi+UnWjhL1wSyqD6fo2Cj4qfHx77rGzCCA==","signature_status":"signed_v1","signed_at":"2026-07-05T07:16:32.299337Z","signed_message":"canonical_sha256_bytes"},"source_id":"2206.10498","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:dd311495da86dad5736aaced04ae5ebdfd239cac5be677e40bf6e9ee3a5f70d8","sha256:3617cc18d4b30a7eaa65387b5efcc55a5f979cf37528d506519d2eabc634f98b"],"state_sha256":"4e58467e37b971dce951c6b4f3ad6f93bb7d8ba5c8e889cc284142991d340c5f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zlUHwh0/aiZdEBQr9Wqpw50gTcqFvwspS/K6yW6FXUz18HOoxhmdhWR+3Zp3aBA8GhOZOgAsO3dhyvUbN5Q/Bg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T18:52:16.579384Z","bundle_sha256":"bc16e5380648bbc970c0d8b83161022769a7ea10b3273dfb4273b2347c3edd01"}}