{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:OQX7BJZRWYOJO4LPUPEHM6FIYV","short_pith_number":"pith:OQX7BJZR","schema_version":"1.0","canonical_sha256":"742ff0a731b61c97716fa3c87678a8c57a98a92cc102af94a3c0477f467e6bde","source":{"kind":"arxiv","id":"2408.16228","version":1},"attestation_state":"computed","paper":{"title":"Policy Adaptation via Language Optimization: Decomposing Tasks for Few-Shot Imitation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Bill Chunyuan Zheng, Kuan Fang, Oier Mees, Sergey Levine, Vivek Myers","submitted_at":"2024-08-29T03:03:35Z","abstract_excerpt":"Learned language-conditioned robot policies often struggle to effectively adapt to new real-world tasks even when pre-trained across a diverse set of instructions. We propose a novel approach for few-shot adaptation to unseen tasks that exploits the semantic understanding of task decomposition provided by vision-language models (VLMs). Our method, Policy Adaptation via Language Optimization (PALO), combines a handful of demonstrations of a task with proposed language decompositions sampled from a VLM to quickly enable rapid nonparametric adaptation, avoiding the need for a larger fine-tuning d"},"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":"2408.16228","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2024-08-29T03:03:35Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"a0029520af0d191e62e7c8a92b1f83d7a454a4d374fba840eac25c86592f4d3b","abstract_canon_sha256":"7abe52763cdce02cedd428ea27cd603f7d4b2bacd5d152b3d5ef983dd9e86c87"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:58:23.939038Z","signature_b64":"XdLdKd5jT7iXgb0J6M0PhcqcMXMIued+JZ2O75219ssPTTwUQmKTKHwc6ol3m43IfGquOJZrjLIY9EITK7OWDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"742ff0a731b61c97716fa3c87678a8c57a98a92cc102af94a3c0477f467e6bde","last_reissued_at":"2026-07-05T09:58:23.938596Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:58:23.938596Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Policy Adaptation via Language Optimization: Decomposing Tasks for Few-Shot Imitation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Bill Chunyuan Zheng, Kuan Fang, Oier Mees, Sergey Levine, Vivek Myers","submitted_at":"2024-08-29T03:03:35Z","abstract_excerpt":"Learned language-conditioned robot policies often struggle to effectively adapt to new real-world tasks even when pre-trained across a diverse set of instructions. We propose a novel approach for few-shot adaptation to unseen tasks that exploits the semantic understanding of task decomposition provided by vision-language models (VLMs). Our method, Policy Adaptation via Language Optimization (PALO), combines a handful of demonstrations of a task with proposed language decompositions sampled from a VLM to quickly enable rapid nonparametric adaptation, avoiding the need for a larger fine-tuning d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.16228","kind":"arxiv","version":1},"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/2408.16228/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":"2408.16228","created_at":"2026-07-05T09:58:23.938654+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.16228v1","created_at":"2026-07-05T09:58:23.938654+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.16228","created_at":"2026-07-05T09:58:23.938654+00:00"},{"alias_kind":"pith_short_12","alias_value":"OQX7BJZRWYOJ","created_at":"2026-07-05T09:58:23.938654+00:00"},{"alias_kind":"pith_short_16","alias_value":"OQX7BJZRWYOJO4LP","created_at":"2026-07-05T09:58:23.938654+00:00"},{"alias_kind":"pith_short_8","alias_value":"OQX7BJZR","created_at":"2026-07-05T09:58:23.938654+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.12299","citing_title":"Learning What to Say to Your VLA: Mostly Harmless Vision Language Action Model Steering","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06461","citing_title":"Flow-based Policy Adaptation without Policy Updates","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23249","citing_title":"BridgeACT: Bridging Human Demonstrations to Robot Actions via Unified Tool-Target Affordances","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV","json":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV.json","graph_json":"https://pith.science/api/pith-number/OQX7BJZRWYOJO4LPUPEHM6FIYV/graph.json","events_json":"https://pith.science/api/pith-number/OQX7BJZRWYOJO4LPUPEHM6FIYV/events.json","paper":"https://pith.science/paper/OQX7BJZR"},"agent_actions":{"view_html":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV","download_json":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV.json","view_paper":"https://pith.science/paper/OQX7BJZR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.16228&json=true","fetch_graph":"https://pith.science/api/pith-number/OQX7BJZRWYOJO4LPUPEHM6FIYV/graph.json","fetch_events":"https://pith.science/api/pith-number/OQX7BJZRWYOJO4LPUPEHM6FIYV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV/action/storage_attestation","attest_author":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV/action/author_attestation","sign_citation":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV/action/citation_signature","submit_replication":"https://pith.science/pith/OQX7BJZRWYOJO4LPUPEHM6FIYV/action/replication_record"}},"created_at":"2026-07-05T09:58:23.938654+00:00","updated_at":"2026-07-05T09:58:23.938654+00:00"}