{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:XVCRH6RYITHXJIQETYAN7HUVPK","short_pith_number":"pith:XVCRH6RY","schema_version":"1.0","canonical_sha256":"bd4513fa3844cf74a2049e00df9e957abb7abf095e6db67de19d8f8c8767e517","source":{"kind":"arxiv","id":"2508.15669","version":1},"attestation_state":"computed","paper":{"title":"Exploiting Policy Idling for Dexterous Manipulation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Annie S. Chen, Annie Xie, Antonia Bronars, Dushyant Rao, Maria Bauza, Markus Wulfmeier, Nicolas Heess, Oliver Groth, Philemon Brakel, Sandy Huang","submitted_at":"2025-08-21T15:52:45Z","abstract_excerpt":"Learning-based methods for dexterous manipulation have made notable progress in recent years. However, learned policies often still lack reliability and exhibit limited robustness to important factors of variation. One failure pattern that can be observed across many settings is that policies idle, i.e. they cease to move beyond a small region of states when they reach certain states. This policy idling is often a reflection of the training data. For instance, it can occur when the data contains small actions in areas where the robot needs to perform high-precision motions, e.g., when preparin"},"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":"2508.15669","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2025-08-21T15:52:45Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3254d4eb5f9638422a531034c7c908749e66f39c73e10bf427ea175aa92ec473","abstract_canon_sha256":"602fb6c121feade5eaa983a699cd66a7c0bafbb6b8198c2a6db7b1666043ccd1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:57:16.683855Z","signature_b64":"kmkcWQ3NskR02lkYLyMWBrHv+YOFKzKOwRu2QByHxlE07uQDfOgnkhY4SWxHRTvRd9/DwrGBNRqQxanxZgAhBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bd4513fa3844cf74a2049e00df9e957abb7abf095e6db67de19d8f8c8767e517","last_reissued_at":"2026-07-05T11:57:16.683252Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:57:16.683252Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploiting Policy Idling for Dexterous Manipulation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.RO","authors_text":"Annie S. Chen, Annie Xie, Antonia Bronars, Dushyant Rao, Maria Bauza, Markus Wulfmeier, Nicolas Heess, Oliver Groth, Philemon Brakel, Sandy Huang","submitted_at":"2025-08-21T15:52:45Z","abstract_excerpt":"Learning-based methods for dexterous manipulation have made notable progress in recent years. However, learned policies often still lack reliability and exhibit limited robustness to important factors of variation. One failure pattern that can be observed across many settings is that policies idle, i.e. they cease to move beyond a small region of states when they reach certain states. This policy idling is often a reflection of the training data. For instance, it can occur when the data contains small actions in areas where the robot needs to perform high-precision motions, e.g., when preparin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.15669","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/2508.15669/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":"2508.15669","created_at":"2026-07-05T11:57:16.683322+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.15669v1","created_at":"2026-07-05T11:57:16.683322+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.15669","created_at":"2026-07-05T11:57:16.683322+00:00"},{"alias_kind":"pith_short_12","alias_value":"XVCRH6RYITHX","created_at":"2026-07-05T11:57:16.683322+00:00"},{"alias_kind":"pith_short_16","alias_value":"XVCRH6RYITHXJIQE","created_at":"2026-07-05T11:57:16.683322+00:00"},{"alias_kind":"pith_short_8","alias_value":"XVCRH6RY","created_at":"2026-07-05T11:57:16.683322+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.15674","citing_title":"Bayesian Optimization with Expected Improvement: No Regret and the Choice of Incumbent","ref_index":1,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK","json":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK.json","graph_json":"https://pith.science/api/pith-number/XVCRH6RYITHXJIQETYAN7HUVPK/graph.json","events_json":"https://pith.science/api/pith-number/XVCRH6RYITHXJIQETYAN7HUVPK/events.json","paper":"https://pith.science/paper/XVCRH6RY"},"agent_actions":{"view_html":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK","download_json":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK.json","view_paper":"https://pith.science/paper/XVCRH6RY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.15669&json=true","fetch_graph":"https://pith.science/api/pith-number/XVCRH6RYITHXJIQETYAN7HUVPK/graph.json","fetch_events":"https://pith.science/api/pith-number/XVCRH6RYITHXJIQETYAN7HUVPK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK/action/storage_attestation","attest_author":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK/action/author_attestation","sign_citation":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK/action/citation_signature","submit_replication":"https://pith.science/pith/XVCRH6RYITHXJIQETYAN7HUVPK/action/replication_record"}},"created_at":"2026-07-05T11:57:16.683322+00:00","updated_at":"2026-07-05T11:57:16.683322+00:00"}