{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:IZSYSQ2QJELOBAE4I2OEZX7KBK","short_pith_number":"pith:IZSYSQ2Q","schema_version":"1.0","canonical_sha256":"46658943504916e0809c469c4cdfea0a885dc8dae9f6166a81bd752e7ca2a901","source":{"kind":"arxiv","id":"2309.11235","version":2},"attestation_state":"computed","paper":{"title":"OpenChat: Advancing Open-source Language Models with Mixed-Quality Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guan Wang, Sen Song, Sijie Cheng, Xiangang Li, Xianyuan Zhan, Yang Liu","submitted_at":"2023-09-20T11:54:40Z","abstract_excerpt":"Nowadays, open-source large language models like LLaMA have emerged. Recent developments have incorporated supervised fine-tuning (SFT) and reinforcement learning fine-tuning (RLFT) to align these models with human goals. However, SFT methods treat all training data with mixed quality equally, while RLFT methods require high-quality pairwise or ranking-based preference data. In this study, we present a novel framework, named OpenChat, to advance open-source language models with mixed-quality data. Specifically, we consider the general SFT training data, consisting of a small amount of expert d"},"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":"2309.11235","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-09-20T11:54:40Z","cross_cats_sorted":[],"title_canon_sha256":"c972f7437c8335218bfca9de13a93f48e91ed627482590fbdffc976308fd748a","abstract_canon_sha256":"33101688c6f1e1c2456c19e0d14eb3c149e152bece1b7ba18825c3f9ee240148"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:56:52.538572Z","signature_b64":"UH18Na+ELF4iZrYjbtXs2IF60oCU03oj10EWol18E3HMSBbFPRpjncbf67JK8BU3ix53o4Svn9yvyciwSDq/Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"46658943504916e0809c469c4cdfea0a885dc8dae9f6166a81bd752e7ca2a901","last_reissued_at":"2026-07-05T07:56:52.538124Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:56:52.538124Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"OpenChat: Advancing Open-source Language Models with Mixed-Quality Data","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guan Wang, Sen Song, Sijie Cheng, Xiangang Li, Xianyuan Zhan, Yang Liu","submitted_at":"2023-09-20T11:54:40Z","abstract_excerpt":"Nowadays, open-source large language models like LLaMA have emerged. Recent developments have incorporated supervised fine-tuning (SFT) and reinforcement learning fine-tuning (RLFT) to align these models with human goals. However, SFT methods treat all training data with mixed quality equally, while RLFT methods require high-quality pairwise or ranking-based preference data. In this study, we present a novel framework, named OpenChat, to advance open-source language models with mixed-quality data. Specifically, we consider the general SFT training data, consisting of a small amount of expert d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.11235","kind":"arxiv","version":2},"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/2309.11235/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":"2309.11235","created_at":"2026-07-05T07:56:52.538181+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.11235v2","created_at":"2026-07-05T07:56:52.538181+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.11235","created_at":"2026-07-05T07:56:52.538181+00:00"},{"alias_kind":"pith_short_12","alias_value":"IZSYSQ2QJELO","created_at":"2026-07-05T07:56:52.538181+00:00"},{"alias_kind":"pith_short_16","alias_value":"IZSYSQ2QJELOBAE4","created_at":"2026-07-05T07:56:52.538181+00:00"},{"alias_kind":"pith_short_8","alias_value":"IZSYSQ2Q","created_at":"2026-07-05T07:56:52.538181+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.22676","citing_title":"Skin-Deep: A Geometric Diagnostic for Alignment Fragility in Large Language Model Representations","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00419","citing_title":"Rethinking LLM Ensembling from the Perspective of Mixture Models","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2402.11684","citing_title":"ALLaVA: Harnessing GPT4V-Synthesized Data for Lite Vision-Language Models","ref_index":128,"is_internal_anchor":false},{"citing_arxiv_id":"2401.05561","citing_title":"TrustLLM: Trustworthiness in Large Language Models","ref_index":95,"is_internal_anchor":false},{"citing_arxiv_id":"2404.08144","citing_title":"LLM Agents can Autonomously Exploit One-day Vulnerabilities","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2408.07666","citing_title":"Model Merging in LLMs, MLLMs, and Beyond: Methods, Theories, Applications and Opportunities","ref_index":234,"is_internal_anchor":false},{"citing_arxiv_id":"2601.21351","citing_title":"Analytical Provisioning for Attention-FFN Disaggregated LLM Serving under Stochastic Workloads","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13875","citing_title":"Common-agency Games for Multi-Objective Test-Time Alignment","ref_index":102,"is_internal_anchor":false},{"citing_arxiv_id":"2406.01574","citing_title":"MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00419","citing_title":"Rethinking LLM Ensembling from the Perspective of Mixture Models","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK","json":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK.json","graph_json":"https://pith.science/api/pith-number/IZSYSQ2QJELOBAE4I2OEZX7KBK/graph.json","events_json":"https://pith.science/api/pith-number/IZSYSQ2QJELOBAE4I2OEZX7KBK/events.json","paper":"https://pith.science/paper/IZSYSQ2Q"},"agent_actions":{"view_html":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK","download_json":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK.json","view_paper":"https://pith.science/paper/IZSYSQ2Q","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.11235&json=true","fetch_graph":"https://pith.science/api/pith-number/IZSYSQ2QJELOBAE4I2OEZX7KBK/graph.json","fetch_events":"https://pith.science/api/pith-number/IZSYSQ2QJELOBAE4I2OEZX7KBK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK/action/storage_attestation","attest_author":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK/action/author_attestation","sign_citation":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK/action/citation_signature","submit_replication":"https://pith.science/pith/IZSYSQ2QJELOBAE4I2OEZX7KBK/action/replication_record"}},"created_at":"2026-07-05T07:56:52.538181+00:00","updated_at":"2026-07-05T07:56:52.538181+00:00"}