{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:R2U4FDUROUVZCFF5BNEEKHYNLN","short_pith_number":"pith:R2U4FDUR","canonical_record":{"source":{"id":"2503.23383","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T10:16:25Z","cross_cats_sorted":[],"title_canon_sha256":"e8354db1f4e53bc9a6263fb6951d9dfb53e140b38e81e7487764f7c49b371ca9","abstract_canon_sha256":"e665d2b221fdc9ae39d0eeeb54205e96a3a141fe2bf6efc054967d324adef4ae"},"schema_version":"1.0"},"canonical_sha256":"8ea9c28e91752b9114bd0b48451f0d5b639ff2abc94e924e5782d60be48f645d","source":{"kind":"arxiv","id":"2503.23383","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.23383","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"arxiv_version","alias_value":"2503.23383v1","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.23383","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"pith_short_12","alias_value":"R2U4FDUROUVZ","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"pith_short_16","alias_value":"R2U4FDUROUVZCFF5","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"pith_short_8","alias_value":"R2U4FDUR","created_at":"2026-07-05T10:41:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:R2U4FDUROUVZCFF5BNEEKHYNLN","target":"record","payload":{"canonical_record":{"source":{"id":"2503.23383","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T10:16:25Z","cross_cats_sorted":[],"title_canon_sha256":"e8354db1f4e53bc9a6263fb6951d9dfb53e140b38e81e7487764f7c49b371ca9","abstract_canon_sha256":"e665d2b221fdc9ae39d0eeeb54205e96a3a141fe2bf6efc054967d324adef4ae"},"schema_version":"1.0"},"canonical_sha256":"8ea9c28e91752b9114bd0b48451f0d5b639ff2abc94e924e5782d60be48f645d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:41:35.335932Z","signature_b64":"vsPLGK6OtzKk9pDJ/XDbkQjPjc0P2y8mDPbEqT6tC2/TuCZ5Yl32YGhxMRdLa/g+qhiX5DzGzhnLnyKnzFpIAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8ea9c28e91752b9114bd0b48451f0d5b639ff2abc94e924e5782d60be48f645d","last_reissued_at":"2026-07-05T10:41:35.335436Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:41:35.335436Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.23383","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:41:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yn1jvhQ8cvdzjE/SYH2EpP5wxx47brqjy0mdHtqanrfvGXcQ+Qu1HJVXpHknFA8d5iUhMQL8v3Xdsn6EuFFWAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T23:59:39.385791Z"},"content_sha256":"ac24da00b0505e3ac623d16c1852b022ab6e66895e5bd07469b0e2d2540f9a57","schema_version":"1.0","event_id":"sha256:ac24da00b0505e3ac623d16c1852b022ab6e66895e5bd07469b0e2d2540f9a57"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:R2U4FDUROUVZCFF5BNEEKHYNLN","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"ToRL: Scaling Tool-Integrated RL","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Haoyang Zou, Pengfei Liu, Xuefeng Li","submitted_at":"2025-03-30T10:16:25Z","abstract_excerpt":"We introduce ToRL (Tool-Integrated Reinforcement Learning), a framework for training large language models (LLMs) to autonomously use computational tools via reinforcement learning. Unlike supervised fine-tuning, ToRL allows models to explore and discover optimal strategies for tool use. Experiments with Qwen2.5-Math models show significant improvements: ToRL-7B reaches 43.3\\% accuracy on AIME~24, surpassing reinforcement learning without tool integration by 14\\% and the best existing Tool-Integrated Reasoning (TIR) model by 17\\%. Further analysis reveals emergent behaviors such as strategic t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.23383","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/2503.23383/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:41:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8I6cDujF27wy1foZmMdxZc1IQsSUy8gHyoPlOt4m5EO9Ht6RrEOlxhWfqcvM4Y0WE07HDFX8vnjB0YQ5KuaqDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T23:59:39.386286Z"},"content_sha256":"05eaf48d55a4308acb515484007aec4f3a15712fef16bd57a0132b64e05d12a4","schema_version":"1.0","event_id":"sha256:05eaf48d55a4308acb515484007aec4f3a15712fef16bd57a0132b64e05d12a4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/R2U4FDUROUVZCFF5BNEEKHYNLN/bundle.json","state_url":"https://pith.science/pith/R2U4FDUROUVZCFF5BNEEKHYNLN/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/R2U4FDUROUVZCFF5BNEEKHYNLN/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":"2026-08-01T23:59:39Z","links":{"resolver":"https://pith.science/pith/R2U4FDUROUVZCFF5BNEEKHYNLN","bundle":"https://pith.science/pith/R2U4FDUROUVZCFF5BNEEKHYNLN/bundle.json","state":"https://pith.science/pith/R2U4FDUROUVZCFF5BNEEKHYNLN/state.json","well_known_bundle":"https://pith.science/.well-known/pith/R2U4FDUROUVZCFF5BNEEKHYNLN/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:R2U4FDUROUVZCFF5BNEEKHYNLN","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":"e665d2b221fdc9ae39d0eeeb54205e96a3a141fe2bf6efc054967d324adef4ae","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T10:16:25Z","title_canon_sha256":"e8354db1f4e53bc9a6263fb6951d9dfb53e140b38e81e7487764f7c49b371ca9"},"schema_version":"1.0","source":{"id":"2503.23383","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.23383","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"arxiv_version","alias_value":"2503.23383v1","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.23383","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"pith_short_12","alias_value":"R2U4FDUROUVZ","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"pith_short_16","alias_value":"R2U4FDUROUVZCFF5","created_at":"2026-07-05T10:41:35Z"},{"alias_kind":"pith_short_8","alias_value":"R2U4FDUR","created_at":"2026-07-05T10:41:35Z"}],"graph_snapshots":[{"event_id":"sha256:05eaf48d55a4308acb515484007aec4f3a15712fef16bd57a0132b64e05d12a4","target":"graph","created_at":"2026-07-05T10:41:35Z","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":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2503.23383/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce ToRL (Tool-Integrated Reinforcement Learning), a framework for training large language models (LLMs) to autonomously use computational tools via reinforcement learning. Unlike supervised fine-tuning, ToRL allows models to explore and discover optimal strategies for tool use. Experiments with Qwen2.5-Math models show significant improvements: ToRL-7B reaches 43.3\\% accuracy on AIME~24, surpassing reinforcement learning without tool integration by 14\\% and the best existing Tool-Integrated Reasoning (TIR) model by 17\\%. Further analysis reveals emergent behaviors such as strategic t","authors_text":"Haoyang Zou, Pengfei Liu, Xuefeng Li","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T10:16:25Z","title":"ToRL: Scaling Tool-Integrated RL"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.23383","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ac24da00b0505e3ac623d16c1852b022ab6e66895e5bd07469b0e2d2540f9a57","target":"record","created_at":"2026-07-05T10:41:35Z","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":"e665d2b221fdc9ae39d0eeeb54205e96a3a141fe2bf6efc054967d324adef4ae","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-30T10:16:25Z","title_canon_sha256":"e8354db1f4e53bc9a6263fb6951d9dfb53e140b38e81e7487764f7c49b371ca9"},"schema_version":"1.0","source":{"id":"2503.23383","kind":"arxiv","version":1}},"canonical_sha256":"8ea9c28e91752b9114bd0b48451f0d5b639ff2abc94e924e5782d60be48f645d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"8ea9c28e91752b9114bd0b48451f0d5b639ff2abc94e924e5782d60be48f645d","first_computed_at":"2026-07-05T10:41:35.335436Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:41:35.335436Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vsPLGK6OtzKk9pDJ/XDbkQjPjc0P2y8mDPbEqT6tC2/TuCZ5Yl32YGhxMRdLa/g+qhiX5DzGzhnLnyKnzFpIAw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:41:35.335932Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.23383","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ac24da00b0505e3ac623d16c1852b022ab6e66895e5bd07469b0e2d2540f9a57","sha256:05eaf48d55a4308acb515484007aec4f3a15712fef16bd57a0132b64e05d12a4"],"state_sha256":"6539358c7063a4925d71caaa93524385372a1bfd42ca00c94d575e0110b63cff"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"/ZWwJQsT8Pz5MMvat5MPAA3ipSuVUFyK7UICTcSTmOLN07ttnOiGHotZiZbmBlCCiJaU73o1qYHBNSPzlidgAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-01T23:59:39.389943Z","bundle_sha256":"d4cf24c46782763775222f8affa551ad9945fe5fa7fbbc4c524f5897baf4c0af"}}