{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3CQFX5J2FNML6B5H3S7V6DEMHE","short_pith_number":"pith:3CQFX5J2","schema_version":"1.0","canonical_sha256":"d8a05bf53a2b58bf07a7dcbf5f0c8c3936ee5c2619c89b25e83eb93f14238fe6","source":{"kind":"arxiv","id":"2405.04798","version":3},"attestation_state":"computed","paper":{"title":"From LLMs to Actions: Latent Codes as Bridges in Hierarchical Robot Control","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.RO","authors_text":"Aravind Rajeswaran, Philipp Wu, Pieter Abbeel, Yide Shentu","submitted_at":"2024-05-08T04:14:06Z","abstract_excerpt":"Hierarchical control for robotics has long been plagued by the need to have a well defined interface layer to communicate between high-level task planners and low-level policies. With the advent of LLMs, language has been emerging as a prospective interface layer. However, this has several limitations. Not all tasks can be decomposed into steps that are easily expressible in natural language (e.g. performing a dance routine). Further, it makes end-to-end finetuning on embodied data challenging due to domain shift and catastrophic forgetting. We introduce our method -- Learnable Latent Codes as"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2405.04798","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.RO","submitted_at":"2024-05-08T04:14:06Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"09806f29b40bca9b3a305f98aff4f0be319fb982f261823c609c270bf55b1882","abstract_canon_sha256":"fccac80a0baf50d24c976c45d8670f4bdea65ff7db3cdf90de6c26c67246a3e4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:33:14.549571Z","signature_b64":"e/CnTnWDKb5/JS6nTN9lyTcb0DI9fQChrA4Zm3NctyXu7iBCa2kPfYq+ByAmcRT02DnSmFJw3ZYI5etLJKruBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d8a05bf53a2b58bf07a7dcbf5f0c8c3936ee5c2619c89b25e83eb93f14238fe6","last_reissued_at":"2026-07-05T11:33:14.549070Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:33:14.549070Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From LLMs to Actions: Latent Codes as Bridges in Hierarchical Robot Control","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.RO","authors_text":"Aravind Rajeswaran, Philipp Wu, Pieter Abbeel, Yide Shentu","submitted_at":"2024-05-08T04:14:06Z","abstract_excerpt":"Hierarchical control for robotics has long been plagued by the need to have a well defined interface layer to communicate between high-level task planners and low-level policies. With the advent of LLMs, language has been emerging as a prospective interface layer. However, this has several limitations. Not all tasks can be decomposed into steps that are easily expressible in natural language (e.g. performing a dance routine). Further, it makes end-to-end finetuning on embodied data challenging due to domain shift and catastrophic forgetting. We introduce our method -- Learnable Latent Codes as"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.04798","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/2405.04798/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2405.04798","created_at":"2026-07-05T11:33:14.549121+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.04798v3","created_at":"2026-07-05T11:33:14.549121+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.04798","created_at":"2026-07-05T11:33:14.549121+00:00"},{"alias_kind":"pith_short_12","alias_value":"3CQFX5J2FNML","created_at":"2026-07-05T11:33:14.549121+00:00"},{"alias_kind":"pith_short_16","alias_value":"3CQFX5J2FNML6B5H","created_at":"2026-07-05T11:33:14.549121+00:00"},{"alias_kind":"pith_short_8","alias_value":"3CQFX5J2","created_at":"2026-07-05T11:33:14.549121+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20905","citing_title":"Vesta: A Generalist Embodied Reasoning Model","ref_index":100,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00110","citing_title":"General Covariant Action Modeling: Constructing Generalized Manifolds via Spatio-Temporal Decoupling","ref_index":116,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE","json":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE.json","graph_json":"https://pith.science/api/pith-number/3CQFX5J2FNML6B5H3S7V6DEMHE/graph.json","events_json":"https://pith.science/api/pith-number/3CQFX5J2FNML6B5H3S7V6DEMHE/events.json","paper":"https://pith.science/paper/3CQFX5J2"},"agent_actions":{"view_html":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE","download_json":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE.json","view_paper":"https://pith.science/paper/3CQFX5J2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.04798&json=true","fetch_graph":"https://pith.science/api/pith-number/3CQFX5J2FNML6B5H3S7V6DEMHE/graph.json","fetch_events":"https://pith.science/api/pith-number/3CQFX5J2FNML6B5H3S7V6DEMHE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE/action/storage_attestation","attest_author":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE/action/author_attestation","sign_citation":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE/action/citation_signature","submit_replication":"https://pith.science/pith/3CQFX5J2FNML6B5H3S7V6DEMHE/action/replication_record"}},"created_at":"2026-07-05T11:33:14.549121+00:00","updated_at":"2026-07-05T11:33:14.549121+00:00"}