{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:TWIA2DL6HLKR3TT6N75NFYXGLU","short_pith_number":"pith:TWIA2DL6","schema_version":"1.0","canonical_sha256":"9d900d0d7e3ad51dce7e6ffad2e2e65d087c0d144f9213e0e091f1ccf019a193","source":{"kind":"arxiv","id":"2410.09286","version":1},"attestation_state":"computed","paper":{"title":"Language-Model-Assisted Bi-Level Programming for Reward Learning from Internet Videos","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.RO","authors_text":"Harsh Mahesheka, Wanxin Jin, Zhaoran Wang, Zhixian Xie","submitted_at":"2024-10-11T22:31:39Z","abstract_excerpt":"Learning from Demonstrations, particularly from biological experts like humans and animals, often encounters significant data acquisition challenges. While recent approaches leverage internet videos for learning, they require complex, task-specific pipelines to extract and retarget motion data for the agent. In this work, we introduce a language-model-assisted bi-level programming framework that enables a reinforcement learning agent to directly learn its reward from internet videos, bypassing dedicated data preparation. The framework includes two levels: an upper level where a vision-language"},"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":"2410.09286","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.RO","submitted_at":"2024-10-11T22:31:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"b90878a8b5fb1b794dd68f59fa47049bf8b4b7f2adfb4a86039c6b4dbdd31c88","abstract_canon_sha256":"f9909c34f2195f0b2015ac8603641f5cbe34eec72397d013dbc6315cc76aaddd"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:19:37.305196Z","signature_b64":"jR7haSS7Ssa8hdoKm04j5sHYFGoI1dAXrn6qewanNJ8PIFSz49mg7yKg8m45HcIiHA+go8p5N30jKgAxAYDECg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9d900d0d7e3ad51dce7e6ffad2e2e65d087c0d144f9213e0e091f1ccf019a193","last_reissued_at":"2026-07-05T09:19:37.304716Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:19:37.304716Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Language-Model-Assisted Bi-Level Programming for Reward Learning from Internet Videos","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.RO","authors_text":"Harsh Mahesheka, Wanxin Jin, Zhaoran Wang, Zhixian Xie","submitted_at":"2024-10-11T22:31:39Z","abstract_excerpt":"Learning from Demonstrations, particularly from biological experts like humans and animals, often encounters significant data acquisition challenges. While recent approaches leverage internet videos for learning, they require complex, task-specific pipelines to extract and retarget motion data for the agent. In this work, we introduce a language-model-assisted bi-level programming framework that enables a reinforcement learning agent to directly learn its reward from internet videos, bypassing dedicated data preparation. The framework includes two levels: an upper level where a vision-language"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.09286","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/2410.09286/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":"2410.09286","created_at":"2026-07-05T09:19:37.304773+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.09286v1","created_at":"2026-07-05T09:19:37.304773+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.09286","created_at":"2026-07-05T09:19:37.304773+00:00"},{"alias_kind":"pith_short_12","alias_value":"TWIA2DL6HLKR","created_at":"2026-07-05T09:19:37.304773+00:00"},{"alias_kind":"pith_short_16","alias_value":"TWIA2DL6HLKR3TT6","created_at":"2026-07-05T09:19:37.304773+00:00"},{"alias_kind":"pith_short_8","alias_value":"TWIA2DL6","created_at":"2026-07-05T09:19:37.304773+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.01721","citing_title":"CoRe: Combined Rewards with Vision-Language Model Feedback for Preference-Aligned Reinforcement Learning","ref_index":107,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU","json":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU.json","graph_json":"https://pith.science/api/pith-number/TWIA2DL6HLKR3TT6N75NFYXGLU/graph.json","events_json":"https://pith.science/api/pith-number/TWIA2DL6HLKR3TT6N75NFYXGLU/events.json","paper":"https://pith.science/paper/TWIA2DL6"},"agent_actions":{"view_html":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU","download_json":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU.json","view_paper":"https://pith.science/paper/TWIA2DL6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.09286&json=true","fetch_graph":"https://pith.science/api/pith-number/TWIA2DL6HLKR3TT6N75NFYXGLU/graph.json","fetch_events":"https://pith.science/api/pith-number/TWIA2DL6HLKR3TT6N75NFYXGLU/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU/action/storage_attestation","attest_author":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU/action/author_attestation","sign_citation":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU/action/citation_signature","submit_replication":"https://pith.science/pith/TWIA2DL6HLKR3TT6N75NFYXGLU/action/replication_record"}},"created_at":"2026-07-05T09:19:37.304773+00:00","updated_at":"2026-07-05T09:19:37.304773+00:00"}