{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:WPRT65JIPNJROKLY2GQDPV5O2O","short_pith_number":"pith:WPRT65JI","schema_version":"1.0","canonical_sha256":"b3e33f75287b53172978d1a037d7aed38a2925bc54326f18b3fad08343ee3469","source":{"kind":"arxiv","id":"2412.12953","version":1},"attestation_state":"computed","paper":{"title":"Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.RO"],"primary_cat":"cs.LG","authors_text":"Jyothish Pari, Moritz Reuss, Pulkit Agrawal, Rudolf Lioutikov","submitted_at":"2024-12-17T14:34:51Z","abstract_excerpt":"Diffusion Policies have become widely used in Imitation Learning, offering several appealing properties, such as generating multimodal and discontinuous behavior. As models are becoming larger to capture more complex capabilities, their computational demands increase, as shown by recent scaling laws. Therefore, continuing with the current architectures will present a computational roadblock. To address this gap, we propose Mixture-of-Denoising Experts (MoDE) as a novel policy for Imitation Learning. MoDE surpasses current state-of-the-art Transformer-based Diffusion Policies while enabling par"},"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":"2412.12953","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-17T14:34:51Z","cross_cats_sorted":["cs.RO"],"title_canon_sha256":"a460f6c92a0c99241172a5f90e7247f72c8247037878c529d39974e287a14e21","abstract_canon_sha256":"e0bf0b45280f200c6ebc65376d60f7966cc2546359aa59eb6c0cc90fa9bfd171"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:50:23.232907Z","signature_b64":"jxZEDECP41U65yF3Q97xqB4Bi5/fvj95TWBlu+b/RBj+R98Ole4dKObhdsITJq10PWKWEzZBVePy8Bth7gJSBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b3e33f75287b53172978d1a037d7aed38a2925bc54326f18b3fad08343ee3469","last_reissued_at":"2026-07-05T09:50:23.232378Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:50:23.232378Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficient Diffusion Transformer Policies with Mixture of Expert Denoisers for Multitask Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.RO"],"primary_cat":"cs.LG","authors_text":"Jyothish Pari, Moritz Reuss, Pulkit Agrawal, Rudolf Lioutikov","submitted_at":"2024-12-17T14:34:51Z","abstract_excerpt":"Diffusion Policies have become widely used in Imitation Learning, offering several appealing properties, such as generating multimodal and discontinuous behavior. As models are becoming larger to capture more complex capabilities, their computational demands increase, as shown by recent scaling laws. Therefore, continuing with the current architectures will present a computational roadblock. To address this gap, we propose Mixture-of-Denoising Experts (MoDE) as a novel policy for Imitation Learning. MoDE surpasses current state-of-the-art Transformer-based Diffusion Policies while enabling par"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.12953","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/2412.12953/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":"2412.12953","created_at":"2026-07-05T09:50:23.232446+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.12953v1","created_at":"2026-07-05T09:50:23.232446+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.12953","created_at":"2026-07-05T09:50:23.232446+00:00"},{"alias_kind":"pith_short_12","alias_value":"WPRT65JIPNJR","created_at":"2026-07-05T09:50:23.232446+00:00"},{"alias_kind":"pith_short_16","alias_value":"WPRT65JIPNJROKLY","created_at":"2026-07-05T09:50:23.232446+00:00"},{"alias_kind":"pith_short_8","alias_value":"WPRT65JI","created_at":"2026-07-05T09:50:23.232446+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.20856","citing_title":"DISC: Decoupling Instruction from State-Conditioned Control via Policy Generation","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2511.18085","citing_title":"Continually Evolving Skill Knowledge in Vision Language Action Model","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2601.21971","citing_title":"Supervised Mixture-of-Experts for Surgical Grasping and Retraction","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.03404","citing_title":"Diffusion Policy with Bayesian Expert Selection for Active Multi-Target Tracking","ref_index":23,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O","json":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O.json","graph_json":"https://pith.science/api/pith-number/WPRT65JIPNJROKLY2GQDPV5O2O/graph.json","events_json":"https://pith.science/api/pith-number/WPRT65JIPNJROKLY2GQDPV5O2O/events.json","paper":"https://pith.science/paper/WPRT65JI"},"agent_actions":{"view_html":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O","download_json":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O.json","view_paper":"https://pith.science/paper/WPRT65JI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.12953&json=true","fetch_graph":"https://pith.science/api/pith-number/WPRT65JIPNJROKLY2GQDPV5O2O/graph.json","fetch_events":"https://pith.science/api/pith-number/WPRT65JIPNJROKLY2GQDPV5O2O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O/action/storage_attestation","attest_author":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O/action/author_attestation","sign_citation":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O/action/citation_signature","submit_replication":"https://pith.science/pith/WPRT65JIPNJROKLY2GQDPV5O2O/action/replication_record"}},"created_at":"2026-07-05T09:50:23.232446+00:00","updated_at":"2026-07-05T09:50:23.232446+00:00"}