{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:4TLK2PF52D52STGBV7C6P5XLSK","short_pith_number":"pith:4TLK2PF5","schema_version":"1.0","canonical_sha256":"e4d6ad3cbdd0fba94cc1afc5e7f6eb92834c6f9a3e20524b98eab437544693c8","source":{"kind":"arxiv","id":"2505.14505","version":1},"attestation_state":"computed","paper":{"title":"ModRWKV: Transformer Multimodality in Linear Time","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jiale Kang, Jiang Rui, Qingyu Yin, Weile Li, Zening Lu, Zhouran Ji, Ziyin Yue","submitted_at":"2025-05-20T15:34:36Z","abstract_excerpt":"Currently, most multimodal studies are based on large language models (LLMs) with quadratic-complexity Transformer architectures. While linear models like RNNs enjoy low inference costs, their application has been largely limited to the text-only modality. This work explores the capabilities of modern RNN architectures in multimodal contexts. We propose ModRWKV-a decoupled multimodal framework built upon the RWKV7 architecture as its LLM backbone-which achieves multi-source information fusion through dynamically adaptable heterogeneous modality encoders. We designed the multimodal modules in M"},"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":"2505.14505","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-20T15:34:36Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"23010d8c3002fa0df3f93fedb9759463d1cd3755639149a25056e8096bea74cb","abstract_canon_sha256":"c0fce5f68104b46f9e3c3f162c4e35aa06a37711dd73088ff0077537edd55364"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:06:14.744059Z","signature_b64":"e9l++ZOONj2tz5xumgA0v0rrwcG7BsMECASIY9eODSjfWYpBVaFA/v4Zn32nUz4jtASnj011eZqrTSYRLHA+BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e4d6ad3cbdd0fba94cc1afc5e7f6eb92834c6f9a3e20524b98eab437544693c8","last_reissued_at":"2026-07-05T11:06:14.743550Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:06:14.743550Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ModRWKV: Transformer Multimodality in Linear Time","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Jiale Kang, Jiang Rui, Qingyu Yin, Weile Li, Zening Lu, Zhouran Ji, Ziyin Yue","submitted_at":"2025-05-20T15:34:36Z","abstract_excerpt":"Currently, most multimodal studies are based on large language models (LLMs) with quadratic-complexity Transformer architectures. While linear models like RNNs enjoy low inference costs, their application has been largely limited to the text-only modality. This work explores the capabilities of modern RNN architectures in multimodal contexts. We propose ModRWKV-a decoupled multimodal framework built upon the RWKV7 architecture as its LLM backbone-which achieves multi-source information fusion through dynamically adaptable heterogeneous modality encoders. We designed the multimodal modules in M"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.14505","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/2505.14505/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":"2505.14505","created_at":"2026-07-05T11:06:14.743615+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.14505v1","created_at":"2026-07-05T11:06:14.743615+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.14505","created_at":"2026-07-05T11:06:14.743615+00:00"},{"alias_kind":"pith_short_12","alias_value":"4TLK2PF52D52","created_at":"2026-07-05T11:06:14.743615+00:00"},{"alias_kind":"pith_short_16","alias_value":"4TLK2PF52D52STGB","created_at":"2026-07-05T11:06:14.743615+00:00"},{"alias_kind":"pith_short_8","alias_value":"4TLK2PF5","created_at":"2026-07-05T11:06:14.743615+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK","json":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK.json","graph_json":"https://pith.science/api/pith-number/4TLK2PF52D52STGBV7C6P5XLSK/graph.json","events_json":"https://pith.science/api/pith-number/4TLK2PF52D52STGBV7C6P5XLSK/events.json","paper":"https://pith.science/paper/4TLK2PF5"},"agent_actions":{"view_html":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK","download_json":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK.json","view_paper":"https://pith.science/paper/4TLK2PF5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.14505&json=true","fetch_graph":"https://pith.science/api/pith-number/4TLK2PF52D52STGBV7C6P5XLSK/graph.json","fetch_events":"https://pith.science/api/pith-number/4TLK2PF52D52STGBV7C6P5XLSK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK/action/storage_attestation","attest_author":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK/action/author_attestation","sign_citation":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK/action/citation_signature","submit_replication":"https://pith.science/pith/4TLK2PF52D52STGBV7C6P5XLSK/action/replication_record"}},"created_at":"2026-07-05T11:06:14.743615+00:00","updated_at":"2026-07-05T11:06:14.743615+00:00"}