{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:ZYQFF5SARDW4SGOCW2R7236AAQ","short_pith_number":"pith:ZYQFF5SA","canonical_record":{"source":{"id":"2505.04588","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-07T17:30:22Z","cross_cats_sorted":[],"title_canon_sha256":"da79235b764bba152c0a246fa0c89a521fcd544162d904d39cd880bd9dfefac2","abstract_canon_sha256":"24ee6e81d6ad32332790766812db89532d98c9ec58455280211a714f2719674f"},"schema_version":"1.0"},"canonical_sha256":"ce2052f64088edc919c2b6a3fd6fc00436fd02c0c7b01a2e6d496d72e771c351","source":{"kind":"arxiv","id":"2505.04588","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.04588","created_at":"2026-05-17T23:38:13Z"},{"alias_kind":"arxiv_version","alias_value":"2505.04588v2","created_at":"2026-05-17T23:38:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.04588","created_at":"2026-05-17T23:38:13Z"},{"alias_kind":"pith_short_12","alias_value":"ZYQFF5SARDW4","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"ZYQFF5SARDW4SGOC","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"ZYQFF5SA","created_at":"2026-05-18T12:33:37Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:ZYQFF5SARDW4SGOCW2R7236AAQ","target":"record","payload":{"canonical_record":{"source":{"id":"2505.04588","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-07T17:30:22Z","cross_cats_sorted":[],"title_canon_sha256":"da79235b764bba152c0a246fa0c89a521fcd544162d904d39cd880bd9dfefac2","abstract_canon_sha256":"24ee6e81d6ad32332790766812db89532d98c9ec58455280211a714f2719674f"},"schema_version":"1.0"},"canonical_sha256":"ce2052f64088edc919c2b6a3fd6fc00436fd02c0c7b01a2e6d496d72e771c351","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:13.493633Z","signature_b64":"ZKS3gN5euFI8wZoamQWccgnW9YHVAcTd4dH5rMEY2q/S9iOPWpITHNHtklmw2rwD3PY6+H1QwPBe/dZwCuMuAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ce2052f64088edc919c2b6a3fd6fc00436fd02c0c7b01a2e6d496d72e771c351","last_reissued_at":"2026-05-17T23:38:13.492969Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:13.492969Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.04588","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:38:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KoAhkXZZDJGMkLqnLWg6MSY4Uj2cjfneNa4IJhTdURSVpPbdXPV9lQpdsf3nDfkNB0StXFeNNJpxxdkBsD7LBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T08:01:00.847391Z"},"content_sha256":"5ab8c43ac08182ba2843a5574d4971fe332465cda21a79550374ce5491b856a4","schema_version":"1.0","event_id":"sha256:5ab8c43ac08182ba2843a5574d4971fe332465cda21a79550374ce5491b856a4"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:ZYQFF5SARDW4SGOCW2R7236AAQ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ZeroSearch: Incentivize the Search Capability of LLMs without Searching","license":"http://creativecommons.org/licenses/by/4.0/","headline":"A fine-tuned retrieval module with degrading document quality trains LLMs to match or beat real search engines via RL without live API calls.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fei Huang, Hao Sun, Jiayan Guo, Jingren Zhou, Pengjun Xie, Xuanbo Fan, Yan Zhang, Yingyan Hou, Yong Jiang, Zile Qiao","submitted_at":"2025-05-07T17:30:22Z","abstract_excerpt":"Effective information searching is essential for enhancing the reasoning and generation capabilities of large language models (LLMs). Recent research has explored using reinforcement learning (RL) to improve LLMs' search capabilities by interacting with live search engines in real-world environments. While these approaches show promising results, they face two major challenges: (1) Uncontrolled Document Quality: The quality of documents returned by search engines is often unpredictable, introducing noise and instability into the training process. (2) Prohibitively High API Costs: RL training r"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"a 7B retrieval module achieves comparable performance to the real search engine, while a 14B retrieval module even surpasses it.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That progressively degrading the quality of documents generated by the fine-tuned retrieval module during curriculum rollouts will reliably elicit and improve the main model's reasoning ability in a manner that transfers to real search engine use.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ZeroSearch simulates search engine interactions via supervised fine-tuning of a retrieval module and curriculum-based RL degradation of document quality, achieving comparable or superior performance to real search engines with 7B and 14B modules.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A fine-tuned retrieval module with degrading document quality trains LLMs to match or beat real search engines via RL without live API calls.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"63add1840a8bfb4e65103ed3716ca65c32a188ef6dac7d9a221af46be07a96d6"},"source":{"id":"2505.04588","kind":"arxiv","version":2},"verdict":{"id":"46b1b71d-e3fc-488f-a232-b6ac842ccf38","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T17:38:01.454336Z","strongest_claim":"a 7B retrieval module achieves comparable performance to the real search engine, while a 14B retrieval module even surpasses it.","one_line_summary":"ZeroSearch simulates search engine interactions via supervised fine-tuning of a retrieval module and curriculum-based RL degradation of document quality, achieving comparable or superior performance to real search engines with 7B and 14B modules.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That progressively degrading the quality of documents generated by the fine-tuned retrieval module during curriculum rollouts will reliably elicit and improve the main model's reasoning ability in a manner that transfers to real search engine use.","pith_extraction_headline":"A fine-tuned retrieval module with degrading document quality trains LLMs to match or beat real search engines via RL without live API calls."},"references":{"count":49,"sample":[{"doi":"","year":2023,"title":"A. Asai, Z. Wu, Y . Wang, A. Sil, and H. Hajishirzi. Self-rag: Learning to retrieve, generate, and critique through self-reflection. In The Twelfth International Conference on Learning Representations","work_id":"996ab5c8-ba86-4ab8-b5a4-d75515b5b93c","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"https://arxiv.org/abs/2212.08037","work_id":"3d95359c-7270-48aa-a841-070fb1259b87","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2022,"title":"PaLM: Scaling Language Modeling with Pathways","work_id":"a94f3ef7-2c49-4445-93fe-6ec16aafd966","ref_index":3,"cited_arxiv_id":"2204.02311","is_internal_anchor":true},{"doi":"","year":2024,"title":"The Llama 3 Herd of Models","work_id":"1549a635-88af-4ac1-acfe-51ae7bb53345","ref_index":4,"cited_arxiv_id":"2407.21783","is_internal_anchor":true},{"doi":"","year":2025,"title":"W. Feng, C. Hao, Y . Zhang, J. Song, and H. Wang. Airrag: Activating intrinsic reasoning for retrieval augmented generation via tree-based search. arXiv preprint arXiv:2501.10053, 2025","work_id":"8ef0adb9-684b-4d73-928f-bdf4b8717433","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":49,"snapshot_sha256":"3465e1606eaf41ddc0a675c0597744397e0d3f77a78bdcaa7e9116d26706f4b3","internal_anchors":17},"formal_canon":{"evidence_count":2,"snapshot_sha256":"de7721bd6c3a6a2054a47c0f129ad8be4f90be2cb4a7846abd161fbd1ac5bbeb"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"verdict_id":"46b1b71d-e3fc-488f-a232-b6ac842ccf38"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:38:13Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IuCzTbXAZCPBsacF27wXBTROreCfESOp8SYWSlATWxWQG29TXMDlfK6yvnGyN5WewmL0wQPd3V25+kdYTA21Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T08:01:00.848443Z"},"content_sha256":"978e62a22caded35f76c1bb5b06f3dafe0361973be97872435e5b2318cb69490","schema_version":"1.0","event_id":"sha256:978e62a22caded35f76c1bb5b06f3dafe0361973be97872435e5b2318cb69490"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZYQFF5SARDW4SGOCW2R7236AAQ/bundle.json","state_url":"https://pith.science/pith/ZYQFF5SARDW4SGOCW2R7236AAQ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZYQFF5SARDW4SGOCW2R7236AAQ/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-25T08:01:00Z","links":{"resolver":"https://pith.science/pith/ZYQFF5SARDW4SGOCW2R7236AAQ","bundle":"https://pith.science/pith/ZYQFF5SARDW4SGOCW2R7236AAQ/bundle.json","state":"https://pith.science/pith/ZYQFF5SARDW4SGOCW2R7236AAQ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZYQFF5SARDW4SGOCW2R7236AAQ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:ZYQFF5SARDW4SGOCW2R7236AAQ","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"24ee6e81d6ad32332790766812db89532d98c9ec58455280211a714f2719674f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-07T17:30:22Z","title_canon_sha256":"da79235b764bba152c0a246fa0c89a521fcd544162d904d39cd880bd9dfefac2"},"schema_version":"1.0","source":{"id":"2505.04588","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.04588","created_at":"2026-05-17T23:38:13Z"},{"alias_kind":"arxiv_version","alias_value":"2505.04588v2","created_at":"2026-05-17T23:38:13Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.04588","created_at":"2026-05-17T23:38:13Z"},{"alias_kind":"pith_short_12","alias_value":"ZYQFF5SARDW4","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_16","alias_value":"ZYQFF5SARDW4SGOC","created_at":"2026-05-18T12:33:37Z"},{"alias_kind":"pith_short_8","alias_value":"ZYQFF5SA","created_at":"2026-05-18T12:33:37Z"}],"graph_snapshots":[{"event_id":"sha256:978e62a22caded35f76c1bb5b06f3dafe0361973be97872435e5b2318cb69490","target":"graph","created_at":"2026-05-17T23:38:13Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":4,"items":[{"attestation":"unclaimed","claim_id":"C1","kind":"strongest_claim","source":"verdict.strongest_claim","status":"machine_extracted","text":"a 7B retrieval module achieves comparable performance to the real search engine, while a 14B retrieval module even surpasses it."},{"attestation":"unclaimed","claim_id":"C2","kind":"weakest_assumption","source":"verdict.weakest_assumption","status":"machine_extracted","text":"That progressively degrading the quality of documents generated by the fine-tuned retrieval module during curriculum rollouts will reliably elicit and improve the main model's reasoning ability in a manner that transfers to real search engine use."},{"attestation":"unclaimed","claim_id":"C3","kind":"one_line_summary","source":"verdict.one_line_summary","status":"machine_extracted","text":"ZeroSearch simulates search engine interactions via supervised fine-tuning of a retrieval module and curriculum-based RL degradation of document quality, achieving comparable or superior performance to real search engines with 7B and 14B modules."},{"attestation":"unclaimed","claim_id":"C4","kind":"headline","source":"verdict.pith_extraction.headline","status":"machine_extracted","text":"A fine-tuned retrieval module with degrading document quality trains LLMs to match or beat real search engines via RL without live API calls."}],"snapshot_sha256":"63add1840a8bfb4e65103ed3716ca65c32a188ef6dac7d9a221af46be07a96d6"},"formal_canon":{"evidence_count":2,"snapshot_sha256":"de7721bd6c3a6a2054a47c0f129ad8be4f90be2cb4a7846abd161fbd1ac5bbeb"},"paper":{"abstract_excerpt":"Effective information searching is essential for enhancing the reasoning and generation capabilities of large language models (LLMs). Recent research has explored using reinforcement learning (RL) to improve LLMs' search capabilities by interacting with live search engines in real-world environments. While these approaches show promising results, they face two major challenges: (1) Uncontrolled Document Quality: The quality of documents returned by search engines is often unpredictable, introducing noise and instability into the training process. (2) Prohibitively High API Costs: RL training r","authors_text":"Fei Huang, Hao Sun, Jiayan Guo, Jingren Zhou, Pengjun Xie, Xuanbo Fan, Yan Zhang, Yingyan Hou, Yong Jiang, Zile Qiao","cross_cats":[],"headline":"A fine-tuned retrieval module with degrading document quality trains LLMs to match or beat real search engines via RL without live API calls.","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-07T17:30:22Z","title":"ZeroSearch: Incentivize the Search Capability of LLMs without Searching"},"references":{"count":49,"internal_anchors":17,"resolved_work":49,"sample":[{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":1,"title":"A. Asai, Z. Wu, Y . Wang, A. Sil, and H. Hajishirzi. Self-rag: Learning to retrieve, generate, and critique through self-reflection. In The Twelfth International Conference on Learning Representations","work_id":"996ab5c8-ba86-4ab8-b5a4-d75515b5b93c","year":2023},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":2,"title":"https://arxiv.org/abs/2212.08037","work_id":"3d95359c-7270-48aa-a841-070fb1259b87","year":2022},{"cited_arxiv_id":"2204.02311","doi":"","is_internal_anchor":true,"ref_index":3,"title":"PaLM: Scaling Language Modeling with Pathways","work_id":"a94f3ef7-2c49-4445-93fe-6ec16aafd966","year":2022},{"cited_arxiv_id":"2407.21783","doi":"","is_internal_anchor":true,"ref_index":4,"title":"The Llama 3 Herd of Models","work_id":"1549a635-88af-4ac1-acfe-51ae7bb53345","year":2024},{"cited_arxiv_id":"","doi":"","is_internal_anchor":false,"ref_index":5,"title":"W. Feng, C. Hao, Y . Zhang, J. Song, and H. Wang. Airrag: Activating intrinsic reasoning for retrieval augmented generation via tree-based search. arXiv preprint arXiv:2501.10053, 2025","work_id":"8ef0adb9-684b-4d73-928f-bdf4b8717433","year":2025}],"snapshot_sha256":"3465e1606eaf41ddc0a675c0597744397e0d3f77a78bdcaa7e9116d26706f4b3"},"source":{"id":"2505.04588","kind":"arxiv","version":2},"verdict":{"created_at":"2026-05-17T17:38:01.454336Z","id":"46b1b71d-e3fc-488f-a232-b6ac842ccf38","model_set":{"reader":"grok-4.3"},"one_line_summary":"ZeroSearch simulates search engine interactions via supervised fine-tuning of a retrieval module and curriculum-based RL degradation of document quality, achieving comparable or superior performance to real search engines with 7B and 14B modules.","pipeline_version":"pith-pipeline@v0.9.0","pith_extraction_headline":"A fine-tuned retrieval module with degrading document quality trains LLMs to match or beat real search engines via RL without live API calls.","strongest_claim":"a 7B retrieval module achieves comparable performance to the real search engine, while a 14B retrieval module even surpasses it.","weakest_assumption":"That progressively degrading the quality of documents generated by the fine-tuned retrieval module during curriculum rollouts will reliably elicit and improve the main model's reasoning ability in a manner that transfers to real search engine use."}},"verdict_id":"46b1b71d-e3fc-488f-a232-b6ac842ccf38"}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:5ab8c43ac08182ba2843a5574d4971fe332465cda21a79550374ce5491b856a4","target":"record","created_at":"2026-05-17T23:38:13Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"24ee6e81d6ad32332790766812db89532d98c9ec58455280211a714f2719674f","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-07T17:30:22Z","title_canon_sha256":"da79235b764bba152c0a246fa0c89a521fcd544162d904d39cd880bd9dfefac2"},"schema_version":"1.0","source":{"id":"2505.04588","kind":"arxiv","version":2}},"canonical_sha256":"ce2052f64088edc919c2b6a3fd6fc00436fd02c0c7b01a2e6d496d72e771c351","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ce2052f64088edc919c2b6a3fd6fc00436fd02c0c7b01a2e6d496d72e771c351","first_computed_at":"2026-05-17T23:38:13.492969Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:38:13.492969Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ZKS3gN5euFI8wZoamQWccgnW9YHVAcTd4dH5rMEY2q/S9iOPWpITHNHtklmw2rwD3PY6+H1QwPBe/dZwCuMuAw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:38:13.493633Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.04588","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:5ab8c43ac08182ba2843a5574d4971fe332465cda21a79550374ce5491b856a4","sha256:978e62a22caded35f76c1bb5b06f3dafe0361973be97872435e5b2318cb69490"],"state_sha256":"ba626c62b6c820e2ceb49bfab47a96017d0f8adc023b35a357bb0b08d04e2899"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YOHTzyI80JW0vm/zBqDJbXceIVf3F5dIGC5aQKnO65cx1E+4BpHdpjUFyq/M/Wye/3a7glzi4cbCxjsqFRWpDw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T08:01:00.852385Z","bundle_sha256":"4b3f1b8ffe3ac4ec6b2a91b6a63ce54b48cfce264f6ae11afd08aac61affaed8"}}