{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:F7D3V4QPWP7OWH3FGCYFNPWK2X","short_pith_number":"pith:F7D3V4QP","schema_version":"1.0","canonical_sha256":"2fc7baf20fb3feeb1f6530b056becad5c649aa674869b745dd8e0f955258dbef","source":{"kind":"arxiv","id":"2602.17547","version":3},"attestation_state":"computed","paper":{"title":"KLong: Training LLM Agent for Extremely Long-horizon Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"KLong shows that trajectory-splitting SFT followed by progressive RL lets a 106B agent outperform a 1T model on extremely long-horizon tasks.","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Bryan Hooi, Flood Sung, Jiaheng Zhang, Xinlong Yang, Yanhao Li, Yibo Miao, Yingwei Ma, Yuchong Xie, Yue Liu, Zhiyuan Hu","submitted_at":"2026-02-19T17:01:08Z","abstract_excerpt":"This paper introduces KLong, an open-source LLM agent trained to solve extremely long-horizon tasks. The principle is to first cold-start the model via trajectory-splitting SFT, then scale it via progressive RL training. Specifically, we first activate basic agentic abilities of a base model with a comprehensive SFT recipe. Then, we introduce Research-Factory, an automated pipeline that generates high-quality training data by collecting research papers and constructing evaluation rubrics. Using this pipeline, we build thousands of long-horizon trajectories distilled from Claude 4.5 Sonnet (Thi"},"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":"2602.17547","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-02-19T17:01:08Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"2c3764cb9431ee8c37b47e285be808e3e1a10e56063d48b7f700fdbdbef3690f","abstract_canon_sha256":"9d11ab0048ddf714e0550a92ef28b9b600db89111931cbb6ba79e1f77f1fc3d2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-10T12:16:10.605544Z","signature_b64":"gsSP9hxGmSu5MY2mxrxmgy4trcu1WQ1vrz/HGudaQppHf3NbP9TRb8tofZOczgH81DqpeTMAPtnpPnm+4bzbDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2fc7baf20fb3feeb1f6530b056becad5c649aa674869b745dd8e0f955258dbef","last_reissued_at":"2026-06-10T12:16:10.602892Z","signature_status":"signed_v1","first_computed_at":"2026-06-10T12:16:10.602892Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"KLong: Training LLM Agent for Extremely Long-horizon Tasks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"KLong shows that trajectory-splitting SFT followed by progressive RL lets a 106B agent outperform a 1T model on extremely long-horizon tasks.","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Bryan Hooi, Flood Sung, Jiaheng Zhang, Xinlong Yang, Yanhao Li, Yibo Miao, Yingwei Ma, Yuchong Xie, Yue Liu, Zhiyuan Hu","submitted_at":"2026-02-19T17:01:08Z","abstract_excerpt":"This paper introduces KLong, an open-source LLM agent trained to solve extremely long-horizon tasks. The principle is to first cold-start the model via trajectory-splitting SFT, then scale it via progressive RL training. Specifically, we first activate basic agentic abilities of a base model with a comprehensive SFT recipe. Then, we introduce Research-Factory, an automated pipeline that generates high-quality training data by collecting research papers and constructing evaluation rubrics. Using this pipeline, we build thousands of long-horizon trajectories distilled from Claude 4.5 Sonnet (Thi"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"our proposed KLong (106B) surpasses Kimi K2 Thinking (1T) by 11.28% on PaperBench, and the performance improvement generalizes to other coding benchmarks like SWE-bench Verified and MLE-bench.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That trajectories distilled from Claude 4.5 Sonnet via the Research-Factory pipeline are high-quality and that trajectory-splitting SFT plus progressive RL stages preserve sufficient context and capability for true long-horizon generalization without hidden failure modes.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"KLong trains a 106B LLM agent using trajectory-splitting SFT and progressive RL to outperform a 1T model by 11.28% on PaperBench with generalization to coding benchmarks.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"KLong shows that trajectory-splitting SFT followed by progressive RL lets a 106B agent outperform a 1T model on extremely long-horizon tasks.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"672e21a8ba139cfd1517a415d54751d0de2d8bc6c4efe0145f7f084dda03212d"},"source":{"id":"2602.17547","kind":"arxiv","version":3},"verdict":{"id":"31e98931-e61b-4679-8934-ce8d62549855","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T20:45:41.123186Z","strongest_claim":"our proposed KLong (106B) surpasses Kimi K2 Thinking (1T) by 11.28% on PaperBench, and the performance improvement generalizes to other coding benchmarks like SWE-bench Verified and MLE-bench.","one_line_summary":"KLong trains a 106B LLM agent using trajectory-splitting SFT and progressive RL to outperform a 1T model by 11.28% on PaperBench with generalization to coding benchmarks.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That trajectories distilled from Claude 4.5 Sonnet via the Research-Factory pipeline are high-quality and that trajectory-splitting SFT plus progressive RL stages preserve sufficient context and capability for true long-horizon generalization without hidden failure modes.","pith_extraction_headline":"KLong shows that trajectory-splitting SFT followed by progressive RL lets a 106B agent outperform a 1T model on extremely long-horizon tasks."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2602.17547/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":"2602.17547","created_at":"2026-06-10T12:16:10.603013+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.17547v3","created_at":"2026-06-10T12:16:10.603013+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.17547","created_at":"2026-06-10T12:16:10.603013+00:00"},{"alias_kind":"pith_short_12","alias_value":"F7D3V4QPWP7O","created_at":"2026-06-10T12:16:10.603013+00:00"},{"alias_kind":"pith_short_16","alias_value":"F7D3V4QPWP7OWH3F","created_at":"2026-06-10T12:16:10.603013+00:00"},{"alias_kind":"pith_short_8","alias_value":"F7D3V4QP","created_at":"2026-06-10T12:16:10.603013+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/F7D3V4QPWP7OWH3FGCYFNPWK2X","json":"https://pith.science/pith/F7D3V4QPWP7OWH3FGCYFNPWK2X.json","graph_json":"https://pith.science/api/pith-number/F7D3V4QPWP7OWH3FGCYFNPWK2X/graph.json","events_json":"https://pith.science/api/pith-number/F7D3V4QPWP7OWH3FGCYFNPWK2X/events.json","paper":"https://pith.science/paper/F7D3V4QP"},"agent_actions":{"view_html":"https://pith.science/pith/F7D3V4QPWP7OWH3FGCYFNPWK2X","download_json":"https://pith.science/pith/F7D3V4QPWP7OWH3FGCYFNPWK2X.json","view_paper":"https://pith.science/paper/F7D3V4QP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.17547&json=true","fetch_graph":"https://pith.science/api/pith-number/F7D3V4QPWP7OWH3FGCYFNPWK2X/graph.json","fetch_events":"https://pith.science/api/pith-number/F7D3V4QPWP7OWH3FGCYFNPWK2X/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/F7D3V4QPWP7OWH3FGCYFNPWK2X/action/timestamp_anchor","attest_storage":"https://pith.science/pith/F7D3V4QPWP7OWH3FGCYFNPWK2X/action/storage_attestation","attest_author":"https://pith.science/pith/F7D3V4QPWP7OWH3FGCYFNPWK2X/action/author_attestation","sign_citation":"https://pith.science/pith/F7D3V4QPWP7OWH3FGCYFNPWK2X/action/citation_signature","submit_replication":"https://pith.science/pith/F7D3V4QPWP7OWH3FGCYFNPWK2X/action/replication_record"}},"created_at":"2026-06-10T12:16:10.603013+00:00","updated_at":"2026-06-10T12:16:10.603013+00:00"}