{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:TY37JXRPT62UALQKSWCGT73QF4","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":"f457de8436d8cc6a5a0b374e163fb99bb2d3af0d4c27abdd4fe96f46fa1dc717","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-02-13T02:28:57Z","title_canon_sha256":"6315accc1ed6edee39d6fa856d1ec69ff15edf4b53a73ce2014651694f1ce900"},"schema_version":"1.0","source":{"id":"2402.08178","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.08178","created_at":"2026-07-05T07:44:28Z"},{"alias_kind":"arxiv_version","alias_value":"2402.08178v1","created_at":"2026-07-05T07:44:28Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.08178","created_at":"2026-07-05T07:44:28Z"},{"alias_kind":"pith_short_12","alias_value":"TY37JXRPT62U","created_at":"2026-07-05T07:44:28Z"},{"alias_kind":"pith_short_16","alias_value":"TY37JXRPT62UALQK","created_at":"2026-07-05T07:44:28Z"},{"alias_kind":"pith_short_8","alias_value":"TY37JXRP","created_at":"2026-07-05T07:44:28Z"}],"graph_snapshots":[{"event_id":"sha256:2179a21380197a3a9503af5ebd774371d4f965367114b534b59be695301f3d33","target":"graph","created_at":"2026-07-05T07:44:28Z","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/2402.08178/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models (LLMs) have recently received considerable attention as alternative solutions for task planning. However, comparing the performance of language-oriented task planners becomes difficult, and there exists a dearth of detailed exploration regarding the effects of various factors such as pre-trained model selection and prompt construction. To address this, we propose a benchmark system for automatically quantifying performance of task planning for home-service embodied agents. Task planners are tested on two pairs of datasets and simulators: 1) ALFRED and AI2-THOR, 2) an exte","authors_text":"Hyobin Ong, Jaehong Kim, Jae-Woo Choi, Minsu Jang, Youngwoo Yoon","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-02-13T02:28:57Z","title":"LoTa-Bench: Benchmarking Language-oriented Task Planners for Embodied Agents"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.08178","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:e517470328928c7b8c8135c671855199fbd31ff885e0074231d5012197f6473a","target":"record","created_at":"2026-07-05T07:44:28Z","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":"f457de8436d8cc6a5a0b374e163fb99bb2d3af0d4c27abdd4fe96f46fa1dc717","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-02-13T02:28:57Z","title_canon_sha256":"6315accc1ed6edee39d6fa856d1ec69ff15edf4b53a73ce2014651694f1ce900"},"schema_version":"1.0","source":{"id":"2402.08178","kind":"arxiv","version":1}},"canonical_sha256":"9e37f4de2f9fb5402e0a958469ff702f2818bcabb1700fc5022a4932647eaa26","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9e37f4de2f9fb5402e0a958469ff702f2818bcabb1700fc5022a4932647eaa26","first_computed_at":"2026-07-05T07:44:28.228258Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:44:28.228258Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"010VlcqSEowa+0XRFZ8qlTOmyHrK8/uKYC+6HyQubPjH14WyPlwWUgpbwThCLEMPTEBkfEQYS3t7bxgRH11MBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:44:28.228714Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.08178","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e517470328928c7b8c8135c671855199fbd31ff885e0074231d5012197f6473a","sha256:2179a21380197a3a9503af5ebd774371d4f965367114b534b59be695301f3d33"],"state_sha256":"ac46c1041ecdd24d02940e9c7146f4608d9bf236b123aa3474b4ca590636ae20"}