{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:RXRGM7AQY4XBAXEMLTJWH4YL7N","short_pith_number":"pith:RXRGM7AQ","schema_version":"1.0","canonical_sha256":"8de2667c10c72e105c8c5cd363f30bfb7bf9f17d7b87b15d538cb6b46b096fdc","source":{"kind":"arxiv","id":"2507.02833","version":3},"attestation_state":"computed","paper":{"title":"Generalizing Verifiable Instruction Following","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hamish Ivison, Hannaneh Hajishirzi, Nathan Lambert, Pradeep Dasigi, Saumya Malik, Shengyi Huang, Valentina Pyatkin, Victoria Graf","submitted_at":"2025-07-03T17:44:33Z","abstract_excerpt":"A crucial factor for successful human and AI interaction is the ability of language models or chatbots to follow human instructions precisely. A common feature of instructions are output constraints like ``only answer with yes or no\" or ``mention the word `abrakadabra' at least 3 times\" that the user adds to craft a more useful answer. Even today's strongest models struggle with fulfilling such constraints. We find that most models strongly overfit on a small set of verifiable constraints from the benchmarks that test these abilities, a skill called precise instruction following, and are not a"},"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":"2507.02833","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-07-03T17:44:33Z","cross_cats_sorted":[],"title_canon_sha256":"860dbe3b7fa3cd22521baa295ddce749c7d16f121ab4f6371566ac88638a9818","abstract_canon_sha256":"9be34a2529413970716ce321ee50b55a8341d85bd97f9c88b3eab41c03c3d8ec"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:49.194365Z","signature_b64":"/Dr9xDbGCWmU67Vy9WqR7PbIErH1a7JVNCfj2Xm+ChHYsvBzGW1eUzeDk60pI+T2cpihrJ6vcF83YENkYrX1CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8de2667c10c72e105c8c5cd363f30bfb7bf9f17d7b87b15d538cb6b46b096fdc","last_reissued_at":"2026-05-17T23:38:49.193712Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:49.193712Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generalizing Verifiable Instruction Following","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hamish Ivison, Hannaneh Hajishirzi, Nathan Lambert, Pradeep Dasigi, Saumya Malik, Shengyi Huang, Valentina Pyatkin, Victoria Graf","submitted_at":"2025-07-03T17:44:33Z","abstract_excerpt":"A crucial factor for successful human and AI interaction is the ability of language models or chatbots to follow human instructions precisely. A common feature of instructions are output constraints like ``only answer with yes or no\" or ``mention the word `abrakadabra' at least 3 times\" that the user adds to craft a more useful answer. Even today's strongest models struggle with fulfilling such constraints. We find that most models strongly overfit on a small set of verifiable constraints from the benchmarks that test these abilities, a skill called precise instruction following, and are not a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.02833","kind":"arxiv","version":3},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"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":"2507.02833","created_at":"2026-05-17T23:38:49.193807+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.02833v3","created_at":"2026-05-17T23:38:49.193807+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.02833","created_at":"2026-05-17T23:38:49.193807+00:00"},{"alias_kind":"pith_short_12","alias_value":"RXRGM7AQY4XB","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"RXRGM7AQY4XBAXEM","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"RXRGM7AQ","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":26,"internal_anchor_count":7,"sample":[{"citing_arxiv_id":"2509.21319","citing_title":"RLBFF: Binary Flexible Feedback to bridge between Human Feedback & Verifiable Rewards","ref_index":31,"is_internal_anchor":true},{"citing_arxiv_id":"2605.20382","citing_title":"Do as I Say, Not as I Do: Instruction-Induction Conflict in LLMs","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2605.18643","citing_title":"Post-Trained MoE Can Skip Half Experts via Self-Distillation","ref_index":16,"is_internal_anchor":true},{"citing_arxiv_id":"2605.19276","citing_title":"OpenCompass: A Universal Evaluation Platform for Large Language Models","ref_index":14,"is_internal_anchor":true},{"citing_arxiv_id":"2509.21882","citing_title":"Position: The Hidden Costs and Measurement Gaps of Reinforcement Learning with Verifiable Rewards","ref_index":28,"is_internal_anchor":true},{"citing_arxiv_id":"2509.26388","citing_title":"Game-Time: Evaluating Temporal Dynamics in Spoken Language Models","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2512.20856","citing_title":"NVIDIA Nemotron 3: Efficient and Open Intelligence","ref_index":151,"is_internal_anchor":true},{"citing_arxiv_id":"2604.16322","citing_title":"Steerable Instruction Following Coding Data Synthesis with Actor-Parametric Schema Co-Evolution","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02315","citing_title":"Beyond the Assistant Turn: User Turn Generation as a Probe of Interaction Awareness in Language Models","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02795","citing_title":"Rubrics to Tokens: Bridging Response-level Rubrics and Token-level Rewards in Instruction Following Tasks","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12227","citing_title":"Combining On-Policy Optimization and Distillation for Long-Context Reasoning in Large Language Models","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08520","citing_title":"FlashEvolve: Accelerating Agent Self-Evolution with Asynchronous Stage Orchestration","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10913","citing_title":"Shepherd: A Runtime Substrate Empowering Meta-Agents with a Formalized Execution Trace","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08639","citing_title":"ReLibra: Routing-Replay-Guided Load Balancing for MoE Training in Reinforcement Learning","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06213","citing_title":"Beyond Fixed Benchmarks and Worst-Case Attacks: Dynamic Boundary Evaluation for Language Models","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.22937","citing_title":"AutoPyVerifier: Learning Compact Executable Verifiers for Large Language Model Outputs","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12312","citing_title":"CompliBench: Benchmarking LLM Judges for Compliance Violation Detection in Dialogue Systems","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09443","citing_title":"Many-Tier Instruction Hierarchy in LLM Agents","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07182","citing_title":"Star Elastic: Many-in-One Reasoning LLMs with Efficient Budget Control","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07461","citing_title":"Think-with-Rubrics: From External Evaluator to Internal Reasoning Guidance","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07465","citing_title":"SEIF: Self-Evolving Reinforcement Learning for Instruction Following","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02913","citing_title":"Generate, Filter, Control, Replay: A Comprehensive Survey of Rollout Strategies for LLM Reinforcement Learning","ref_index":91,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15602","citing_title":"GroupDPO: Memory efficient Group-wise Direct Preference Optimization","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15804","citing_title":"Qwen3.5-Omni Technical Report","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15804","citing_title":"Qwen3.5-Omni Technical Report","ref_index":32,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N","json":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N.json","graph_json":"https://pith.science/api/pith-number/RXRGM7AQY4XBAXEMLTJWH4YL7N/graph.json","events_json":"https://pith.science/api/pith-number/RXRGM7AQY4XBAXEMLTJWH4YL7N/events.json","paper":"https://pith.science/paper/RXRGM7AQ"},"agent_actions":{"view_html":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N","download_json":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N.json","view_paper":"https://pith.science/paper/RXRGM7AQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.02833&json=true","fetch_graph":"https://pith.science/api/pith-number/RXRGM7AQY4XBAXEMLTJWH4YL7N/graph.json","fetch_events":"https://pith.science/api/pith-number/RXRGM7AQY4XBAXEMLTJWH4YL7N/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N/action/storage_attestation","attest_author":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N/action/author_attestation","sign_citation":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N/action/citation_signature","submit_replication":"https://pith.science/pith/RXRGM7AQY4XBAXEMLTJWH4YL7N/action/replication_record"}},"created_at":"2026-05-17T23:38:49.193807+00:00","updated_at":"2026-05-17T23:38:49.193807+00:00"}