{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3XU5UE5FLPG6L62FJA6SD6DBX7","short_pith_number":"pith:3XU5UE5F","schema_version":"1.0","canonical_sha256":"dde9da13a55bcde5fb45483d21f861bfe2d0c56f2e436b19a990f4636e9300d8","source":{"kind":"arxiv","id":"2505.12744","version":1},"attestation_state":"computed","paper":{"title":"Incentivizing Multimodal Reasoning in Large Models for Direct Robot Manipulation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chi-Wing Fu, Dong Jing, Jia-Hui Pan, Li Erran Li, Mingyu Ding, Weiliang Tang, Yun-hui Liu, Zhiwu Lu","submitted_at":"2025-05-19T06:00:14Z","abstract_excerpt":"Recent Large Multimodal Models have demonstrated remarkable reasoning capabilities, especially in solving complex mathematical problems and realizing accurate spatial perception. Our key insight is that these emerging abilities can naturally extend to robotic manipulation by enabling LMMs to directly infer the next goal in language via reasoning, rather than relying on a separate action head. However, this paradigm meets two main challenges: i) How to make LMMs understand the spatial action space, and ii) How to fully exploit the reasoning capacity of LMMs in solving these tasks. To tackle the"},"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":"2505.12744","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-05-19T06:00:14Z","cross_cats_sorted":[],"title_canon_sha256":"11335d1d16d3e92a890fd0fbcb2e88aa9af374066c54f9aadd6be830fff344e8","abstract_canon_sha256":"a911639f78ae1db632769c1b3112ddfda824aa4c9f9de9d7e7c99a304e2693f2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:05:09.003140Z","signature_b64":"hHNGBCkGBzjuxF/ASgOxtLE72+n4hVwiULUltWXXHcL/qicYySv6Ayy9OGEhsrplG08UPjmq8l11aUqxnQe2DA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dde9da13a55bcde5fb45483d21f861bfe2d0c56f2e436b19a990f4636e9300d8","last_reissued_at":"2026-07-05T11:05:09.002693Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:05:09.002693Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Incentivizing Multimodal Reasoning in Large Models for Direct Robot Manipulation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chi-Wing Fu, Dong Jing, Jia-Hui Pan, Li Erran Li, Mingyu Ding, Weiliang Tang, Yun-hui Liu, Zhiwu Lu","submitted_at":"2025-05-19T06:00:14Z","abstract_excerpt":"Recent Large Multimodal Models have demonstrated remarkable reasoning capabilities, especially in solving complex mathematical problems and realizing accurate spatial perception. Our key insight is that these emerging abilities can naturally extend to robotic manipulation by enabling LMMs to directly infer the next goal in language via reasoning, rather than relying on a separate action head. However, this paradigm meets two main challenges: i) How to make LMMs understand the spatial action space, and ii) How to fully exploit the reasoning capacity of LMMs in solving these tasks. To tackle the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.12744","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/2505.12744/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":"2505.12744","created_at":"2026-07-05T11:05:09.002749+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.12744v1","created_at":"2026-07-05T11:05:09.002749+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.12744","created_at":"2026-07-05T11:05:09.002749+00:00"},{"alias_kind":"pith_short_12","alias_value":"3XU5UE5FLPG6","created_at":"2026-07-05T11:05:09.002749+00:00"},{"alias_kind":"pith_short_16","alias_value":"3XU5UE5FLPG6L62F","created_at":"2026-07-05T11:05:09.002749+00:00"},{"alias_kind":"pith_short_8","alias_value":"3XU5UE5F","created_at":"2026-07-05T11:05:09.002749+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26095","citing_title":"Learning Action Priors for Cross-embodiment Robot Manipulation","ref_index":41,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7","json":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7.json","graph_json":"https://pith.science/api/pith-number/3XU5UE5FLPG6L62FJA6SD6DBX7/graph.json","events_json":"https://pith.science/api/pith-number/3XU5UE5FLPG6L62FJA6SD6DBX7/events.json","paper":"https://pith.science/paper/3XU5UE5F"},"agent_actions":{"view_html":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7","download_json":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7.json","view_paper":"https://pith.science/paper/3XU5UE5F","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.12744&json=true","fetch_graph":"https://pith.science/api/pith-number/3XU5UE5FLPG6L62FJA6SD6DBX7/graph.json","fetch_events":"https://pith.science/api/pith-number/3XU5UE5FLPG6L62FJA6SD6DBX7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7/action/storage_attestation","attest_author":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7/action/author_attestation","sign_citation":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7/action/citation_signature","submit_replication":"https://pith.science/pith/3XU5UE5FLPG6L62FJA6SD6DBX7/action/replication_record"}},"created_at":"2026-07-05T11:05:09.002749+00:00","updated_at":"2026-07-05T11:05:09.002749+00:00"}