{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VLDJTS2KWXH45PY2GH5OHMYZJF","short_pith_number":"pith:VLDJTS2K","schema_version":"1.0","canonical_sha256":"aac699cb4ab5cfcebf1a31fae3b319497ba2f679fa27eebef15b33a891c909b8","source":{"kind":"arxiv","id":"2506.07905","version":1},"attestation_state":"computed","paper":{"title":"WeThink: Toward General-purpose Vision-Language Reasoning via Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dacheng Yin, Feipeng Ma, Fengyun Rao, Jie Yang, Kang Rong, Ruimao Zhang, Zitian Wang","submitted_at":"2025-06-09T16:20:54Z","abstract_excerpt":"Building on the success of text-based reasoning models like DeepSeek-R1, extending these capabilities to multimodal reasoning holds great promise. While recent works have attempted to adapt DeepSeek-R1-style reinforcement learning (RL) training paradigms to multimodal large language models (MLLM), focusing on domain-specific tasks like math and visual perception, a critical question remains: How can we achieve the general-purpose visual-language reasoning through RL? To address this challenge, we make three key efforts: (1) A novel Scalable Multimodal QA Synthesis pipeline that autonomously ge"},"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":"2506.07905","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-06-09T16:20:54Z","cross_cats_sorted":[],"title_canon_sha256":"9cb173a3e62d053d234522f7329d79bf98cd40a7a246ce741eddc6701ab49958","abstract_canon_sha256":"50687675dbcf468192eeadcd7a9448902d69fbbad1506ff80fbb5745afb09468"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:18:35.211964Z","signature_b64":"OrGSA83OJ1OLHD2wrhkaJlHwqmAmrA3nRBWvbUcUZAln6InIUiz9tkEZH/6yBeiXDNFY7POg7MOUMZfNY57NCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aac699cb4ab5cfcebf1a31fae3b319497ba2f679fa27eebef15b33a891c909b8","last_reissued_at":"2026-07-05T11:18:35.211447Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:18:35.211447Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"WeThink: Toward General-purpose Vision-Language Reasoning via Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dacheng Yin, Feipeng Ma, Fengyun Rao, Jie Yang, Kang Rong, Ruimao Zhang, Zitian Wang","submitted_at":"2025-06-09T16:20:54Z","abstract_excerpt":"Building on the success of text-based reasoning models like DeepSeek-R1, extending these capabilities to multimodal reasoning holds great promise. While recent works have attempted to adapt DeepSeek-R1-style reinforcement learning (RL) training paradigms to multimodal large language models (MLLM), focusing on domain-specific tasks like math and visual perception, a critical question remains: How can we achieve the general-purpose visual-language reasoning through RL? To address this challenge, we make three key efforts: (1) A novel Scalable Multimodal QA Synthesis pipeline that autonomously ge"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.07905","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/2506.07905/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":"2506.07905","created_at":"2026-07-05T11:18:35.211500+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.07905v1","created_at":"2026-07-05T11:18:35.211500+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.07905","created_at":"2026-07-05T11:18:35.211500+00:00"},{"alias_kind":"pith_short_12","alias_value":"VLDJTS2KWXH4","created_at":"2026-07-05T11:18:35.211500+00:00"},{"alias_kind":"pith_short_16","alias_value":"VLDJTS2KWXH45PY2","created_at":"2026-07-05T11:18:35.211500+00:00"},{"alias_kind":"pith_short_8","alias_value":"VLDJTS2K","created_at":"2026-07-05T11:18:35.211500+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23543","citing_title":"VeriEvol: Scaling Multimodal Mathematical Reasoning via Verifiable Evol-Instruct","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20177","citing_title":"From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2509.02547","citing_title":"The Landscape of Agentic Reinforcement Learning for LLMs: A Survey","ref_index":238,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19264","citing_title":"DR-MMSearchAgent: Deepening Reasoning in Multimodal Search Agents","ref_index":75,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF","json":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF.json","graph_json":"https://pith.science/api/pith-number/VLDJTS2KWXH45PY2GH5OHMYZJF/graph.json","events_json":"https://pith.science/api/pith-number/VLDJTS2KWXH45PY2GH5OHMYZJF/events.json","paper":"https://pith.science/paper/VLDJTS2K"},"agent_actions":{"view_html":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF","download_json":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF.json","view_paper":"https://pith.science/paper/VLDJTS2K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.07905&json=true","fetch_graph":"https://pith.science/api/pith-number/VLDJTS2KWXH45PY2GH5OHMYZJF/graph.json","fetch_events":"https://pith.science/api/pith-number/VLDJTS2KWXH45PY2GH5OHMYZJF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF/action/storage_attestation","attest_author":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF/action/author_attestation","sign_citation":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF/action/citation_signature","submit_replication":"https://pith.science/pith/VLDJTS2KWXH45PY2GH5OHMYZJF/action/replication_record"}},"created_at":"2026-07-05T11:18:35.211500+00:00","updated_at":"2026-07-05T11:18:35.211500+00:00"}