{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:IEIAB3JIF4L3TKMQHPACUOBZDY","short_pith_number":"pith:IEIAB3JI","schema_version":"1.0","canonical_sha256":"411000ed282f17b9a9903bc02a38391e3b333200604d2cd7c9bf036f54a3e888","source":{"kind":"arxiv","id":"2412.08442","version":1},"attestation_state":"computed","paper":{"title":"From Multimodal LLMs to Generalist Embodied Agents: Methods and Lessons","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aleksei Timofeev, Alexander Toshev, Andrew Szot, Bogdan Mazoure, Devon Hjelm, Harsh Agrawal, Omar Attia, Zhe Gan, Zsolt Kira","submitted_at":"2024-12-11T15:06:25Z","abstract_excerpt":"We examine the capability of Multimodal Large Language Models (MLLMs) to tackle diverse domains that extend beyond the traditional language and vision tasks these models are typically trained on. Specifically, our focus lies in areas such as Embodied AI, Games, UI Control, and Planning. To this end, we introduce a process of adapting an MLLM to a Generalist Embodied Agent (GEA). GEA is a single unified model capable of grounding itself across these varied domains through a multi-embodiment action tokenizer. GEA is trained with supervised learning on a large dataset of embodied experiences and "},"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":"2412.08442","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-11T15:06:25Z","cross_cats_sorted":[],"title_canon_sha256":"b6b6c3b9649522a269c80efbd8d3f9e49919919e3b0859238919a7fe7fd4482f","abstract_canon_sha256":"0bf2236b72e3b6da2e59a0e83c348137ac6ba9a5996f3e41d63e37383f6ef3f0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:47:50.201965Z","signature_b64":"RZA2G7vR1TTbKyHgLR7yWwTzvr0KBcNPQ5XPRrRGPfGJzhmmROb1lyJ7znPL22+8RSjrnq79kNzWJrw9rFp6DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"411000ed282f17b9a9903bc02a38391e3b333200604d2cd7c9bf036f54a3e888","last_reissued_at":"2026-07-05T09:47:50.201575Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:47:50.201575Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Multimodal LLMs to Generalist Embodied Agents: Methods and Lessons","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Aleksei Timofeev, Alexander Toshev, Andrew Szot, Bogdan Mazoure, Devon Hjelm, Harsh Agrawal, Omar Attia, Zhe Gan, Zsolt Kira","submitted_at":"2024-12-11T15:06:25Z","abstract_excerpt":"We examine the capability of Multimodal Large Language Models (MLLMs) to tackle diverse domains that extend beyond the traditional language and vision tasks these models are typically trained on. Specifically, our focus lies in areas such as Embodied AI, Games, UI Control, and Planning. To this end, we introduce a process of adapting an MLLM to a Generalist Embodied Agent (GEA). GEA is a single unified model capable of grounding itself across these varied domains through a multi-embodiment action tokenizer. GEA is trained with supervised learning on a large dataset of embodied experiences and "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.08442","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/2412.08442/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":"2412.08442","created_at":"2026-07-05T09:47:50.201632+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.08442v1","created_at":"2026-07-05T09:47:50.201632+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.08442","created_at":"2026-07-05T09:47:50.201632+00:00"},{"alias_kind":"pith_short_12","alias_value":"IEIAB3JIF4L3","created_at":"2026-07-05T09:47:50.201632+00:00"},{"alias_kind":"pith_short_16","alias_value":"IEIAB3JIF4L3TKMQ","created_at":"2026-07-05T09:47:50.201632+00:00"},{"alias_kind":"pith_short_8","alias_value":"IEIAB3JI","created_at":"2026-07-05T09:47:50.201632+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20905","citing_title":"Vesta: A Generalist Embodied Reasoning Model","ref_index":106,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00110","citing_title":"General Covariant Action Modeling: Constructing Generalized Manifolds via Spatio-Temporal Decoupling","ref_index":154,"is_internal_anchor":false},{"citing_arxiv_id":"2504.16054","citing_title":"$\\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization","ref_index":75,"is_internal_anchor":false},{"citing_arxiv_id":"2507.16815","citing_title":"ThinkAct: Vision-Language-Action Reasoning via Reinforced Visual Latent Planning","ref_index":44,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY","json":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY.json","graph_json":"https://pith.science/api/pith-number/IEIAB3JIF4L3TKMQHPACUOBZDY/graph.json","events_json":"https://pith.science/api/pith-number/IEIAB3JIF4L3TKMQHPACUOBZDY/events.json","paper":"https://pith.science/paper/IEIAB3JI"},"agent_actions":{"view_html":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY","download_json":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY.json","view_paper":"https://pith.science/paper/IEIAB3JI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.08442&json=true","fetch_graph":"https://pith.science/api/pith-number/IEIAB3JIF4L3TKMQHPACUOBZDY/graph.json","fetch_events":"https://pith.science/api/pith-number/IEIAB3JIF4L3TKMQHPACUOBZDY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY/action/storage_attestation","attest_author":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY/action/author_attestation","sign_citation":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY/action/citation_signature","submit_replication":"https://pith.science/pith/IEIAB3JIF4L3TKMQHPACUOBZDY/action/replication_record"}},"created_at":"2026-07-05T09:47:50.201632+00:00","updated_at":"2026-07-05T09:47:50.201632+00:00"}