{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VCHEE3M72EMLXDP6BS6XRTPM4F","short_pith_number":"pith:VCHEE3M7","schema_version":"1.0","canonical_sha256":"a88e426d9fd118bb8dfe0cbd78cdece17791709132953930893245620c56f97e","source":{"kind":"arxiv","id":"2501.18096","version":1},"attestation_state":"computed","paper":{"title":"LLMs can see and hear without any training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Ishan Misra, Kumar Ashutosh, Rohit Girdhar, Xinlei Chen, Yossi Gandelsman","submitted_at":"2025-01-30T02:16:35Z","abstract_excerpt":"We present MILS: Multimodal Iterative LLM Solver, a surprisingly simple, training-free approach, to imbue multimodal capabilities into your favorite LLM. Leveraging their innate ability to perform multi-step reasoning, MILS prompts the LLM to generate candidate outputs, each of which are scored and fed back iteratively, eventually generating a solution to the task. This enables various applications that typically require training specialized models on task-specific data. In particular, we establish a new state-of-the-art on emergent zero-shot image, video and audio captioning. MILS seamlessly "},"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":"2501.18096","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-01-30T02:16:35Z","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"title_canon_sha256":"bcd6b421fc3e6ccab52b5925012bd75272351053104af96766b0dbd3b3c4148c","abstract_canon_sha256":"b2bb74e343baa3ec5de9937014342f70a91be7f8f8985d577cdcb9758cde64d9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:07:06.972196Z","signature_b64":"tvHKWDdz1J3/tpUKFzOgLhGczJyvwDpZ0posMms9Bx5RSMDiZgkon+qZ1KccpgrJ4CXHP1+ZBzaRFWgW18VRCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a88e426d9fd118bb8dfe0cbd78cdece17791709132953930893245620c56f97e","last_reissued_at":"2026-07-05T10:07:06.971750Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:07:06.971750Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLMs can see and hear without any training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL","cs.LG"],"primary_cat":"cs.CV","authors_text":"Ishan Misra, Kumar Ashutosh, Rohit Girdhar, Xinlei Chen, Yossi Gandelsman","submitted_at":"2025-01-30T02:16:35Z","abstract_excerpt":"We present MILS: Multimodal Iterative LLM Solver, a surprisingly simple, training-free approach, to imbue multimodal capabilities into your favorite LLM. Leveraging their innate ability to perform multi-step reasoning, MILS prompts the LLM to generate candidate outputs, each of which are scored and fed back iteratively, eventually generating a solution to the task. This enables various applications that typically require training specialized models on task-specific data. In particular, we establish a new state-of-the-art on emergent zero-shot image, video and audio captioning. MILS seamlessly "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.18096","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/2501.18096/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":"2501.18096","created_at":"2026-07-05T10:07:06.971807+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.18096v1","created_at":"2026-07-05T10:07:06.971807+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.18096","created_at":"2026-07-05T10:07:06.971807+00:00"},{"alias_kind":"pith_short_12","alias_value":"VCHEE3M72EML","created_at":"2026-07-05T10:07:06.971807+00:00"},{"alias_kind":"pith_short_16","alias_value":"VCHEE3M72EMLXDP6","created_at":"2026-07-05T10:07:06.971807+00:00"},{"alias_kind":"pith_short_8","alias_value":"VCHEE3M7","created_at":"2026-07-05T10:07:06.971807+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2601.00090","citing_title":"It's Never Too Late: Noise Optimization for Collapse Recovery in Trained Diffusion Models","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07427","citing_title":"Personalizing Text-to-Image Generation to Individual Taste","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F","json":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F.json","graph_json":"https://pith.science/api/pith-number/VCHEE3M72EMLXDP6BS6XRTPM4F/graph.json","events_json":"https://pith.science/api/pith-number/VCHEE3M72EMLXDP6BS6XRTPM4F/events.json","paper":"https://pith.science/paper/VCHEE3M7"},"agent_actions":{"view_html":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F","download_json":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F.json","view_paper":"https://pith.science/paper/VCHEE3M7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.18096&json=true","fetch_graph":"https://pith.science/api/pith-number/VCHEE3M72EMLXDP6BS6XRTPM4F/graph.json","fetch_events":"https://pith.science/api/pith-number/VCHEE3M72EMLXDP6BS6XRTPM4F/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F/action/storage_attestation","attest_author":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F/action/author_attestation","sign_citation":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F/action/citation_signature","submit_replication":"https://pith.science/pith/VCHEE3M72EMLXDP6BS6XRTPM4F/action/replication_record"}},"created_at":"2026-07-05T10:07:06.971807+00:00","updated_at":"2026-07-05T10:07:06.971807+00:00"}