{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:VIAW4PMPOHYVVJYYTWBK3JH7F5","short_pith_number":"pith:VIAW4PMP","canonical_record":{"source":{"id":"2605.07066","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-08T00:17:33Z","cross_cats_sorted":[],"title_canon_sha256":"5b529320c8728de9af002b36202f427c6b3259fc19b53fdeb0e86748b126dabe","abstract_canon_sha256":"1542e9f7a33f4ca5981b4c90c26d920c3faa5a066e154bd717d7e16a69cc286a"},"schema_version":"1.0"},"canonical_sha256":"aa016e3d8f71f15aa7189d82ada4ff2f70492113faeb61831650fef2c777bd7e","source":{"kind":"arxiv","id":"2605.07066","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.07066","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"arxiv_version","alias_value":"2605.07066v2","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.07066","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"pith_short_12","alias_value":"VIAW4PMPOHYV","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"pith_short_16","alias_value":"VIAW4PMPOHYVVJYY","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"pith_short_8","alias_value":"VIAW4PMP","created_at":"2026-05-20T01:06:09Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:VIAW4PMPOHYVVJYYTWBK3JH7F5","target":"record","payload":{"canonical_record":{"source":{"id":"2605.07066","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-08T00:17:33Z","cross_cats_sorted":[],"title_canon_sha256":"5b529320c8728de9af002b36202f427c6b3259fc19b53fdeb0e86748b126dabe","abstract_canon_sha256":"1542e9f7a33f4ca5981b4c90c26d920c3faa5a066e154bd717d7e16a69cc286a"},"schema_version":"1.0"},"canonical_sha256":"aa016e3d8f71f15aa7189d82ada4ff2f70492113faeb61831650fef2c777bd7e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:06:09.883264Z","signature_b64":"VrZ87GphDdCZMEI6WOkjWPo3jgxrA5byWigqK3Bzb5IgUY9uaEyDgvGy5QAJVRggQB9SVku8CslhjX/0yKonBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aa016e3d8f71f15aa7189d82ada4ff2f70492113faeb61831650fef2c777bd7e","last_reissued_at":"2026-05-20T01:06:09.882679Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:06:09.882679Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.07066","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T01:06:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KWoPjcaAzxYyY1aZuYBwJEZGYmkvRrPBGSr6+41fxjNEeMmPC440i+egkBIF/ZPRhEHB857cCA7i8ZIQ7Le9Dg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T16:16:06.943668Z"},"content_sha256":"00c49ae4010268ca5f09966f6657c0f273662352dc118572dae5e8535864ee44","schema_version":"1.0","event_id":"sha256:00c49ae4010268ca5f09966f6657c0f273662352dc118572dae5e8535864ee44"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:VIAW4PMPOHYVVJYYTWBK3JH7F5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"2.5-D Decomposition for LLM-Based Spatial Construction","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A 2.5-D decomposition lets LLMs build structures from language instructions by planning only the horizontal plane while a deterministic executor computes vertical placements from column occupancy.","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Li-Jen Chen, Paul Whitten, Sharath Baddam","submitted_at":"2026-05-08T00:17:33Z","abstract_excerpt":"Autonomous systems that build structures from natural-language instructions need reliable spatial reasoning, yet large language models (LLMs) make systematic coordinate errors when generating three-dimensional block placements. We present a neuro-symbolic pipeline based on \\emph{2.5-D decomposition}: the LLM plans in the two-dimensional horizontal plane while a deterministic executor computes all vertical placement from column occupancy, eliminating an entire class of errors. On the Build What I Mean benchmark (160 rounds), GPT-4o-mini with this pipeline achieves 94.6\\% mean structural accurac"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"On the Build What I Mean benchmark (160 rounds), GPT-4o-mini with this pipeline achieves 94.6% mean structural accuracy across 12 independent runs, within 3.0 percentage points of the 97.6% ceiling imposed by architect-agent errors that no builder-side improvement can address. This outperforms both GPT-4o at 90.3% and the best competing system at 76.3%.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That vertical block placements are fully and correctly determined solely by column occupancy without requiring additional spatial reasoning or handling complex inter-block dependencies beyond simple stacking.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"2.5-D decomposition lets LLMs achieve 94.6% structural accuracy on a building benchmark by handling only horizontal planning while a symbolic system manages vertical placements from occupancy.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A 2.5-D decomposition lets LLMs build structures from language instructions by planning only the horizontal plane while a deterministic executor computes vertical placements from column occupancy.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"ce6d46e50058680625448091d3675ec734afa4175bc4a8fc1950f888b6b93851"},"source":{"id":"2605.07066","kind":"arxiv","version":2},"verdict":{"id":"0b1a9de1-ebbc-4e74-9271-2e98c156cbdb","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-11T01:31:26.926216Z","strongest_claim":"On the Build What I Mean benchmark (160 rounds), GPT-4o-mini with this pipeline achieves 94.6% mean structural accuracy across 12 independent runs, within 3.0 percentage points of the 97.6% ceiling imposed by architect-agent errors that no builder-side improvement can address. This outperforms both GPT-4o at 90.3% and the best competing system at 76.3%.","one_line_summary":"2.5-D decomposition lets LLMs achieve 94.6% structural accuracy on a building benchmark by handling only horizontal planning while a symbolic system manages vertical placements from occupancy.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That vertical block placements are fully and correctly determined solely by column occupancy without requiring additional spatial reasoning or handling complex inter-block dependencies beyond simple stacking.","pith_extraction_headline":"A 2.5-D decomposition lets LLMs build structures from language instructions by planning only the horizontal plane while a deterministic executor computes vertical placements from column occupancy."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.07066/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T17:31:18.715353Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T12:09:30.060770Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"9c04d00e8a73e396ca98832b3ba26f9a3eedfe47f9447eb17ff63c5acce6153a"},"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":"0b1a9de1-ebbc-4e74-9271-2e98c156cbdb"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-20T01:06:09Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8H8AxZK764sj/McrOQ5kAt0dmHsuJaI+HLeWp+Te9mlRgNuIm59XHAsFBNVVnITeBebd/v+FxUoihbYmTuaWAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-30T16:16:06.944730Z"},"content_sha256":"c48bbbc917e0738be3b6b4726a0901ffe867d56f68e986e9987420a6fd6f13f3","schema_version":"1.0","event_id":"sha256:c48bbbc917e0738be3b6b4726a0901ffe867d56f68e986e9987420a6fd6f13f3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VIAW4PMPOHYVVJYYTWBK3JH7F5/bundle.json","state_url":"https://pith.science/pith/VIAW4PMPOHYVVJYYTWBK3JH7F5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VIAW4PMPOHYVVJYYTWBK3JH7F5/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":"2026-05-30T16:16:06Z","links":{"resolver":"https://pith.science/pith/VIAW4PMPOHYVVJYYTWBK3JH7F5","bundle":"https://pith.science/pith/VIAW4PMPOHYVVJYYTWBK3JH7F5/bundle.json","state":"https://pith.science/pith/VIAW4PMPOHYVVJYYTWBK3JH7F5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VIAW4PMPOHYVVJYYTWBK3JH7F5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:VIAW4PMPOHYVVJYYTWBK3JH7F5","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":"1542e9f7a33f4ca5981b4c90c26d920c3faa5a066e154bd717d7e16a69cc286a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-08T00:17:33Z","title_canon_sha256":"5b529320c8728de9af002b36202f427c6b3259fc19b53fdeb0e86748b126dabe"},"schema_version":"1.0","source":{"id":"2605.07066","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.07066","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"arxiv_version","alias_value":"2605.07066v2","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.07066","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"pith_short_12","alias_value":"VIAW4PMPOHYV","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"pith_short_16","alias_value":"VIAW4PMPOHYVVJYY","created_at":"2026-05-20T01:06:09Z"},{"alias_kind":"pith_short_8","alias_value":"VIAW4PMP","created_at":"2026-05-20T01:06:09Z"}],"graph_snapshots":[{"event_id":"sha256:c48bbbc917e0738be3b6b4726a0901ffe867d56f68e986e9987420a6fd6f13f3","target":"graph","created_at":"2026-05-20T01:06:09Z","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":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"On the Build What I Mean benchmark (160 rounds), GPT-4o-mini with this pipeline achieves 94.6% mean structural accuracy across 12 independent runs, within 3.0 percentage points of the 97.6% ceiling imposed by architect-agent errors that no builder-side improvement can address. This outperforms both GPT-4o at 90.3% and the best competing system at 76.3%."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That vertical block placements are fully and correctly determined solely by column occupancy without requiring additional spatial reasoning or handling complex inter-block dependencies beyond simple stacking."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"2.5-D decomposition lets LLMs achieve 94.6% structural accuracy on a building benchmark by handling only horizontal planning while a symbolic system manages vertical placements from occupancy."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A 2.5-D decomposition lets LLMs build structures from language instructions by planning only the horizontal plane while a deterministic executor computes vertical placements from column occupancy."}],"snapshot_sha256":"ce6d46e50058680625448091d3675ec734afa4175bc4a8fc1950f888b6b93851"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[{"findings_count":0,"name":"doi_title_agreement","ran_at":"2026-05-19T17:31:18.715353Z","status":"completed","version":"1.0.0"},{"findings_count":0,"name":"doi_compliance","ran_at":"2026-05-19T12:09:30.060770Z","status":"completed","version":"1.0.0"}],"endpoint":"/pith/2605.07066/integrity.json","findings":[],"snapshot_sha256":"9c04d00e8a73e396ca98832b3ba26f9a3eedfe47f9447eb17ff63c5acce6153a","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Autonomous systems that build structures from natural-language instructions need reliable spatial reasoning, yet large language models (LLMs) make systematic coordinate errors when generating three-dimensional block placements. We present a neuro-symbolic pipeline based on \\emph{2.5-D decomposition}: the LLM plans in the two-dimensional horizontal plane while a deterministic executor computes all vertical placement from column occupancy, eliminating an entire class of errors. On the Build What I Mean benchmark (160 rounds), GPT-4o-mini with this pipeline achieves 94.6\\% mean structural accurac","authors_text":"Li-Jen Chen, Paul Whitten, Sharath Baddam","cross_cats":[],"headline":"A 2.5-D decomposition lets LLMs build structures from language instructions by planning only the horizontal plane while a deterministic executor computes vertical placements from column occupancy.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-08T00:17:33Z","title":"2.5-D Decomposition for LLM-Based Spatial Construction"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.07066","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-11T01:31:26.926216Z","id":"0b1a9de1-ebbc-4e74-9271-2e98c156cbdb","model_set":{"reader":"grok-4.3"},"one_line_summary":"2.5-D decomposition lets LLMs achieve 94.6% structural accuracy on a building benchmark by handling only horizontal planning while a symbolic system manages vertical placements from occupancy.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A 2.5-D decomposition lets LLMs build structures from language instructions by planning only the horizontal plane while a deterministic executor computes vertical placements from column occupancy.","strongest_claim":"On the Build What I Mean benchmark (160 rounds), GPT-4o-mini with this pipeline achieves 94.6% mean structural accuracy across 12 independent runs, within 3.0 percentage points of the 97.6% ceiling imposed by architect-agent errors that no builder-side improvement can address. This outperforms both GPT-4o at 90.3% and the best competing system at 76.3%.","weakest_assumption":"That vertical block placements are fully and correctly determined solely by column occupancy without requiring additional spatial reasoning or handling complex inter-block dependencies beyond simple stacking."}},"verdict_id":"0b1a9de1-ebbc-4e74-9271-2e98c156cbdb"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:00c49ae4010268ca5f09966f6657c0f273662352dc118572dae5e8535864ee44","target":"record","created_at":"2026-05-20T01:06:09Z","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":"1542e9f7a33f4ca5981b4c90c26d920c3faa5a066e154bd717d7e16a69cc286a","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-08T00:17:33Z","title_canon_sha256":"5b529320c8728de9af002b36202f427c6b3259fc19b53fdeb0e86748b126dabe"},"schema_version":"1.0","source":{"id":"2605.07066","kind":"arxiv","version":2}},"canonical_sha256":"aa016e3d8f71f15aa7189d82ada4ff2f70492113faeb61831650fef2c777bd7e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"aa016e3d8f71f15aa7189d82ada4ff2f70492113faeb61831650fef2c777bd7e","first_computed_at":"2026-05-20T01:06:09.882679Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T01:06:09.882679Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"VrZ87GphDdCZMEI6WOkjWPo3jgxrA5byWigqK3Bzb5IgUY9uaEyDgvGy5QAJVRggQB9SVku8CslhjX/0yKonBw==","signature_status":"signed_v1","signed_at":"2026-05-20T01:06:09.883264Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.07066","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:00c49ae4010268ca5f09966f6657c0f273662352dc118572dae5e8535864ee44","sha256:c48bbbc917e0738be3b6b4726a0901ffe867d56f68e986e9987420a6fd6f13f3"],"state_sha256":"e9094dd66740aafee93a4ccf3e772ecb3f90c9eaa7b1c026df087d495a4af543"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PJ25elT5Ab33Qfbncy7QhOSB96KNKYWXdL89Nrznn7oIqJAGw++QI4f5NPNlqgJKnfIlR6MFS6fiejYMAhhHBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-30T16:16:06.951022Z","bundle_sha256":"c9b993a2df63fd49dab4eff01d1c2b131aaa23ef0a5dd0b7d3a5c43e5cfbfc31"}}