{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:QSMJBTON7HYSTF5FNUCYM4HMX7","short_pith_number":"pith:QSMJBTON","schema_version":"1.0","canonical_sha256":"849890cdcdf9f12997a56d058670ecbfd5930fbc9e64cf9cfee3aa5bc4dcd078","source":{"kind":"arxiv","id":"2412.14780","version":1},"attestation_state":"computed","paper":{"title":"Disentangling Reasoning Tokens and Boilerplate Tokens For Language Model Fine-tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fuli Feng, Jianxin Ma, Junyang Lin, Yang Zhang, Zhenru Zhang, Ziang Ye","submitted_at":"2024-12-19T12:06:24Z","abstract_excerpt":"When using agent-task datasets to enhance agent capabilities for Large Language Models (LLMs), current methodologies often treat all tokens within a sample equally. However, we argue that tokens serving different roles - specifically, reasoning tokens versus boilerplate tokens (e.g., those governing output format) - differ significantly in importance and learning complexity, necessitating their disentanglement and distinct treatment. To address this, we propose a novel Shuffle-Aware Discriminator (SHAD) for adaptive token discrimination. SHAD classifies tokens by exploiting predictability diff"},"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":"2412.14780","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-12-19T12:06:24Z","cross_cats_sorted":[],"title_canon_sha256":"4b45a0a6b940e9cd1c9701a70d402bc31d2503c5064054cc3b433b1069da6fdc","abstract_canon_sha256":"7e06c6b02da3612d29ec6e8f1d9b844fc7d193648d1bae1153c6b1e7c331e990"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:51:52.704891Z","signature_b64":"QCLLblsujKZ1JVNQgA0ixpVNooXeYbaHi3gliAIpbZA6ryF13CxkKY4G5OQSsqVfb8BcSr5s2CpmHQuQHTa8Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"849890cdcdf9f12997a56d058670ecbfd5930fbc9e64cf9cfee3aa5bc4dcd078","last_reissued_at":"2026-07-05T09:51:52.704421Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:51:52.704421Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Disentangling Reasoning Tokens and Boilerplate Tokens For Language Model Fine-tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Fuli Feng, Jianxin Ma, Junyang Lin, Yang Zhang, Zhenru Zhang, Ziang Ye","submitted_at":"2024-12-19T12:06:24Z","abstract_excerpt":"When using agent-task datasets to enhance agent capabilities for Large Language Models (LLMs), current methodologies often treat all tokens within a sample equally. However, we argue that tokens serving different roles - specifically, reasoning tokens versus boilerplate tokens (e.g., those governing output format) - differ significantly in importance and learning complexity, necessitating their disentanglement and distinct treatment. To address this, we propose a novel Shuffle-Aware Discriminator (SHAD) for adaptive token discrimination. SHAD classifies tokens by exploiting predictability diff"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.14780","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/2412.14780/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":"2412.14780","created_at":"2026-07-05T09:51:52.704473+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.14780v1","created_at":"2026-07-05T09:51:52.704473+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.14780","created_at":"2026-07-05T09:51:52.704473+00:00"},{"alias_kind":"pith_short_12","alias_value":"QSMJBTON7HYS","created_at":"2026-07-05T09:51:52.704473+00:00"},{"alias_kind":"pith_short_16","alias_value":"QSMJBTON7HYSTF5F","created_at":"2026-07-05T09:51:52.704473+00:00"},{"alias_kind":"pith_short_8","alias_value":"QSMJBTON","created_at":"2026-07-05T09:51:52.704473+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/QSMJBTON7HYSTF5FNUCYM4HMX7","json":"https://pith.science/pith/QSMJBTON7HYSTF5FNUCYM4HMX7.json","graph_json":"https://pith.science/api/pith-number/QSMJBTON7HYSTF5FNUCYM4HMX7/graph.json","events_json":"https://pith.science/api/pith-number/QSMJBTON7HYSTF5FNUCYM4HMX7/events.json","paper":"https://pith.science/paper/QSMJBTON"},"agent_actions":{"view_html":"https://pith.science/pith/QSMJBTON7HYSTF5FNUCYM4HMX7","download_json":"https://pith.science/pith/QSMJBTON7HYSTF5FNUCYM4HMX7.json","view_paper":"https://pith.science/paper/QSMJBTON","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.14780&json=true","fetch_graph":"https://pith.science/api/pith-number/QSMJBTON7HYSTF5FNUCYM4HMX7/graph.json","fetch_events":"https://pith.science/api/pith-number/QSMJBTON7HYSTF5FNUCYM4HMX7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QSMJBTON7HYSTF5FNUCYM4HMX7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QSMJBTON7HYSTF5FNUCYM4HMX7/action/storage_attestation","attest_author":"https://pith.science/pith/QSMJBTON7HYSTF5FNUCYM4HMX7/action/author_attestation","sign_citation":"https://pith.science/pith/QSMJBTON7HYSTF5FNUCYM4HMX7/action/citation_signature","submit_replication":"https://pith.science/pith/QSMJBTON7HYSTF5FNUCYM4HMX7/action/replication_record"}},"created_at":"2026-07-05T09:51:52.704473+00:00","updated_at":"2026-07-05T09:51:52.704473+00:00"}