{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:FYN4RFFBVFSWUUUGMZ3R5ZSEP2","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":"92b344e6172175696b4c494f461c8d89070242449cfbc5c528c263a6933d5a7a","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2024-03-04T08:16:11Z","title_canon_sha256":"6ce889699add44e9d8826eff1f6ba9e286a1dd915a3843ad177b00c773c46637"},"schema_version":"1.0","source":{"id":"2403.01823","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.01823","created_at":"2026-05-17T23:38:14Z"},{"alias_kind":"arxiv_version","alias_value":"2403.01823v2","created_at":"2026-05-17T23:38:14Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.01823","created_at":"2026-05-17T23:38:14Z"},{"alias_kind":"pith_short_12","alias_value":"FYN4RFFBVFSW","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"FYN4RFFBVFSWUUUG","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"FYN4RFFB","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:744eebc612735224fcb31edaae6655227a7e1cefabd9468ee6fdc9bd0b6ea24b","target":"graph","created_at":"2026-05-17T23:38:14Z","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":"Our method RT-H builds an action hierarchy using language motions: it first learns to predict language motions, and conditioned on this and the high-level task, it predicts actions, using visual context at all stages."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That fine-grained language motion phrases capture shared low-level structure across semantically diverse tasks sufficiently well that predicting them improves downstream action prediction and enables effective language-based correction."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"RT-H learns robot policies by first predicting language motions as an intermediate representation and then mapping those plus the high-level task to actions, yielding more robust multi-task performance and the ability to learn from language interventions."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"Predicting fine-grained language descriptions of motions first helps robot policies share structure across diverse tasks and accept language corrections."}],"snapshot_sha256":"e4cd9ca7d9f39dc9c49fc7a13fe4adc5f791a2bd56de0da0e2cc8adbec4df1ee"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"37f1e93178923e324fa86c6fca24ba794483d79cbc5af82b0566d27c73a78a7c"},"paper":{"abstract_excerpt":"Language provides a way to break down complex concepts into digestible pieces. Recent works in robot imitation learning use language-conditioned policies that predict actions given visual observations and the high-level task specified in language. These methods leverage the structure of natural language to share data between semantically similar tasks (e.g., \"pick coke can\" and \"pick an apple\") in multi-task datasets. However, as tasks become more semantically diverse (e.g., \"pick coke can\" and \"pour cup\"), sharing data between tasks becomes harder, so learning to map high-level tasks to actio","authors_text":"Debidatta Dwibedi, Dorsa Sadigh, Jonathan Tompson, Pierre Sermanet, Quon Vuong, Suneel Belkhale, Ted Xiao, Tianli Ding, Yevgen Chebotar","cross_cats":["cs.AI"],"headline":"Predicting fine-grained language descriptions of motions first helps robot policies share structure across diverse tasks and accept language corrections.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2024-03-04T08:16:11Z","title":"RT-H: Action Hierarchies Using Language"},"references":{"count":63,"internal_anchors":6,"resolved_work":63,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"Do as i can, not as i say: Grounding language in robotic affordances","work_id":"162aa552-ee2f-4cc9-b1b6-f951beeed62a","year":2023},{"cited_arxiv_id":"","doi":"10.1145/3568162.3578623","is_internal_anchor":false,"ref_index":2,"title":"“No, to the Right","work_id":"ab15835c-e9db-4f47-8325-d798c6f35c30","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":3,"title":"Correcting robot plans with natural language feedback","work_id":"bff9f09b-3261-4eca-be1a-447a04fcbb45","year":null},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":4,"title":"URL https://api.semanticscholar.org/CorpusID: 248085271","work_id":"6d0c3ebd-641a-4481-872f-00df32ae5ec0","year":null},{"cited_arxiv_id":"2307.15818","doi":"","is_internal_anchor":true,"ref_index":5,"title":"RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control","work_id":"ff438a8a-8003-4fae-9131-acd418b3597b","year":2023}],"snapshot_sha256":"aee9348c11220fac061336eef4a2cd7afb74f430ef6ede64bb6495a16cf2f4c5"},"source":{"id":"2403.01823","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-17T06:49:12.759689Z","id":"8b740000-7a5e-4ec6-9195-f24f2cebc662","model_set":{"reader":"grok-4.3"},"one_line_summary":"RT-H learns robot policies by first predicting language motions as an intermediate representation and then mapping those plus the high-level task to actions, yielding more robust multi-task performance and the ability to learn from language interventions.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"Predicting fine-grained language descriptions of motions first helps robot policies share structure across diverse tasks and accept language corrections.","strongest_claim":"Our method RT-H builds an action hierarchy using language motions: it first learns to predict language motions, and conditioned on this and the high-level task, it predicts actions, using visual context at all stages.","weakest_assumption":"That fine-grained language motion phrases capture shared low-level structure across semantically diverse tasks sufficiently well that predicting them improves downstream action prediction and enables effective language-based correction."}},"verdict_id":"8b740000-7a5e-4ec6-9195-f24f2cebc662"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:4b002381ae096647020e7a9d91d13f7009d9e44ebf55ce2e7e48e80509842994","target":"record","created_at":"2026-05-17T23:38:14Z","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":"92b344e6172175696b4c494f461c8d89070242449cfbc5c528c263a6933d5a7a","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2024-03-04T08:16:11Z","title_canon_sha256":"6ce889699add44e9d8826eff1f6ba9e286a1dd915a3843ad177b00c773c46637"},"schema_version":"1.0","source":{"id":"2403.01823","kind":"arxiv","version":2}},"canonical_sha256":"2e1bc894a1a9656a528666771ee6447eb469f837ee05bea1f488a7363f760f38","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2e1bc894a1a9656a528666771ee6447eb469f837ee05bea1f488a7363f760f38","first_computed_at":"2026-05-17T23:38:14.778993Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:14.778993Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"c5vtWYFCcHCub3drw/fdCwVuDuouMd+HOlG9feJXBrxB1sUWyDArYg/S82K5VP8UguOHIMfu/HX4SUmdPuNyCw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:14.779553Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.01823","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4b002381ae096647020e7a9d91d13f7009d9e44ebf55ce2e7e48e80509842994","sha256:744eebc612735224fcb31edaae6655227a7e1cefabd9468ee6fdc9bd0b6ea24b"],"state_sha256":"7fd02ebe0381ff1dcdac86fa33d53b8802ae37d56df59b2fcb9d02a868335a32"}