{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HDS6VKIQFEZPJHCHE4FVVZ2DQI","short_pith_number":"pith:HDS6VKIQ","schema_version":"1.0","canonical_sha256":"38e5eaa9102932f49c47270b5ae74382230dbe62afef41d961383d1998bb6528","source":{"kind":"arxiv","id":"2505.18557","version":2},"attestation_state":"computed","paper":{"title":"TAG-INSTRUCT: Controlled Instruction Complexity Enhancement through Structure-based Augmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guanhua Chen, He Zhu, Junyou Su, Wenjia Zhang, Xingwei He, Yun Chen, Zhiwen Ruan","submitted_at":"2025-05-24T06:51:03Z","abstract_excerpt":"High-quality instruction data is crucial for developing large language models (LLMs), yet existing approaches struggle to effectively control instruction complexity. We present TAG-INSTRUCT, a novel framework that enhances instruction complexity through structured semantic compression and controlled difficulty augmentation. Unlike previous prompt-based methods operating on raw text, TAG-INSTRUCT compresses instructions into a compact tag space and systematically enhances complexity through RL-guided tag expansion. Through extensive experiments, we show that TAG-INSTRUCT outperforms existing in"},"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":"2505.18557","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-24T06:51:03Z","cross_cats_sorted":[],"title_canon_sha256":"831a1cad0ab1b5a095c9d835728cbe1c39ecf6c0be882621e8667c7f0c35e64f","abstract_canon_sha256":"8420feeef92efbc8f600ee0c8addc8d65866f6e2273d48fe433b5a629e11f16a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:14:07.803333Z","signature_b64":"Ei+oGDoEhiK7g2Bqnl/30JJVcqsEQyxIztHwfKC3WQ9MJMqZDtrV+BqmcugfgIRzWSH86yg6OO5V6TVvrgEzBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"38e5eaa9102932f49c47270b5ae74382230dbe62afef41d961383d1998bb6528","last_reissued_at":"2026-07-05T11:14:07.802797Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:14:07.802797Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TAG-INSTRUCT: Controlled Instruction Complexity Enhancement through Structure-based Augmentation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Guanhua Chen, He Zhu, Junyou Su, Wenjia Zhang, Xingwei He, Yun Chen, Zhiwen Ruan","submitted_at":"2025-05-24T06:51:03Z","abstract_excerpt":"High-quality instruction data is crucial for developing large language models (LLMs), yet existing approaches struggle to effectively control instruction complexity. We present TAG-INSTRUCT, a novel framework that enhances instruction complexity through structured semantic compression and controlled difficulty augmentation. Unlike previous prompt-based methods operating on raw text, TAG-INSTRUCT compresses instructions into a compact tag space and systematically enhances complexity through RL-guided tag expansion. Through extensive experiments, we show that TAG-INSTRUCT outperforms existing in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.18557","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/2505.18557/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":"2505.18557","created_at":"2026-07-05T11:14:07.802854+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.18557v2","created_at":"2026-07-05T11:14:07.802854+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.18557","created_at":"2026-07-05T11:14:07.802854+00:00"},{"alias_kind":"pith_short_12","alias_value":"HDS6VKIQFEZP","created_at":"2026-07-05T11:14:07.802854+00:00"},{"alias_kind":"pith_short_16","alias_value":"HDS6VKIQFEZPJHCH","created_at":"2026-07-05T11:14:07.802854+00:00"},{"alias_kind":"pith_short_8","alias_value":"HDS6VKIQ","created_at":"2026-07-05T11:14:07.802854+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/HDS6VKIQFEZPJHCHE4FVVZ2DQI","json":"https://pith.science/pith/HDS6VKIQFEZPJHCHE4FVVZ2DQI.json","graph_json":"https://pith.science/api/pith-number/HDS6VKIQFEZPJHCHE4FVVZ2DQI/graph.json","events_json":"https://pith.science/api/pith-number/HDS6VKIQFEZPJHCHE4FVVZ2DQI/events.json","paper":"https://pith.science/paper/HDS6VKIQ"},"agent_actions":{"view_html":"https://pith.science/pith/HDS6VKIQFEZPJHCHE4FVVZ2DQI","download_json":"https://pith.science/pith/HDS6VKIQFEZPJHCHE4FVVZ2DQI.json","view_paper":"https://pith.science/paper/HDS6VKIQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.18557&json=true","fetch_graph":"https://pith.science/api/pith-number/HDS6VKIQFEZPJHCHE4FVVZ2DQI/graph.json","fetch_events":"https://pith.science/api/pith-number/HDS6VKIQFEZPJHCHE4FVVZ2DQI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HDS6VKIQFEZPJHCHE4FVVZ2DQI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HDS6VKIQFEZPJHCHE4FVVZ2DQI/action/storage_attestation","attest_author":"https://pith.science/pith/HDS6VKIQFEZPJHCHE4FVVZ2DQI/action/author_attestation","sign_citation":"https://pith.science/pith/HDS6VKIQFEZPJHCHE4FVVZ2DQI/action/citation_signature","submit_replication":"https://pith.science/pith/HDS6VKIQFEZPJHCHE4FVVZ2DQI/action/replication_record"}},"created_at":"2026-07-05T11:14:07.802854+00:00","updated_at":"2026-07-05T11:14:07.802854+00:00"}