{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:DFVYWZHKCJEWOUNW3PRJFZK2CE","short_pith_number":"pith:DFVYWZHK","schema_version":"1.0","canonical_sha256":"196b8b64ea12496751b6dbe292e55a11350f3bb2a52a27c1d2031ff26a32402f","source":{"kind":"arxiv","id":"2502.01100","version":2},"attestation_state":"computed","paper":{"title":"ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Ashish Sabharwal, Bill Yuchen Lin, Kyle Richardson, Peter Clark, Radha Poovendran, Ronan Le Bras, Yejin Choi","submitted_at":"2025-02-03T06:44:49Z","abstract_excerpt":"We investigate the logical reasoning capabilities of large language models (LLMs) and their scalability in complex non-monotonic reasoning. To this end, we introduce ZebraLogic, a comprehensive evaluation framework for assessing LLM reasoning performance on logic grid puzzles derived from constraint satisfaction problems (CSPs). ZebraLogic enables the generation of puzzles with controllable and quantifiable complexity, facilitating a systematic study of the scaling limits of models such as Llama, o1 models, and DeepSeek-R1. By encompassing a broad range of search space complexities and diverse"},"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":"2502.01100","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-02-03T06:44:49Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"8218633ce41279d54796d5ecb6e0e45be7dd0da8436105160443a13ebc8a83f7","abstract_canon_sha256":"6501993ff04be18b17060b4eaf78eb088c5c92833afd47e0bce4d26abe9e5b3f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:37:15.554967Z","signature_b64":"NQtcymeAZT4COMAxajTZPh1MkAym0oqUR/D6GsC2Ca9j49/s+UU/0JrMJq36BQnVXB4GH9LD4N4W1b3AoUsMBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"196b8b64ea12496751b6dbe292e55a11350f3bb2a52a27c1d2031ff26a32402f","last_reissued_at":"2026-07-05T11:37:15.554311Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:37:15.554311Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ZebraLogic: On the Scaling Limits of LLMs for Logical Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Ashish Sabharwal, Bill Yuchen Lin, Kyle Richardson, Peter Clark, Radha Poovendran, Ronan Le Bras, Yejin Choi","submitted_at":"2025-02-03T06:44:49Z","abstract_excerpt":"We investigate the logical reasoning capabilities of large language models (LLMs) and their scalability in complex non-monotonic reasoning. To this end, we introduce ZebraLogic, a comprehensive evaluation framework for assessing LLM reasoning performance on logic grid puzzles derived from constraint satisfaction problems (CSPs). ZebraLogic enables the generation of puzzles with controllable and quantifiable complexity, facilitating a systematic study of the scaling limits of models such as Llama, o1 models, and DeepSeek-R1. By encompassing a broad range of search space complexities and diverse"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.01100","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/2502.01100/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":"2502.01100","created_at":"2026-07-05T11:37:15.554405+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.01100v2","created_at":"2026-07-05T11:37:15.554405+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.01100","created_at":"2026-07-05T11:37:15.554405+00:00"},{"alias_kind":"pith_short_12","alias_value":"DFVYWZHKCJEW","created_at":"2026-07-05T11:37:15.554405+00:00"},{"alias_kind":"pith_short_16","alias_value":"DFVYWZHKCJEWOUNW","created_at":"2026-07-05T11:37:15.554405+00:00"},{"alias_kind":"pith_short_8","alias_value":"DFVYWZHK","created_at":"2026-07-05T11:37:15.554405+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":28,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.25178","citing_title":"Transferability for General Reasoning: An Automated Curriculum for Multi-Domain RLVR","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26671","citing_title":"NebulaExp-8B: An Empirical Post-Training Pipeline via Full-Scale Ablation Research","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11470","citing_title":"The Periodic Table of LLM Reasoning: A Structured Survey of Reasoning Paradigms, Methods, and Failure Modes","ref_index":146,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10722","citing_title":"Continual LLM Upcycling: A Predictor-Gated Bank-Wise Sparsity Training Recipe for Dense-to-Sparse LLMs","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2606.08068","citing_title":"DICE: Entropy-Regularized Equilibrium Selection for Stable Multi-Agent LLM Coordination","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01182","citing_title":"CA-BED: Conversation-Aware Bayesian Experimental Design","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25178","citing_title":"Transferability for General Reasoning: An Automated Curriculum for Multi-Domain RLVR","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30626","citing_title":"DOPD: Dual On-policy Distillation","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28742","citing_title":"CORE: Contrastive Reflection Enables Rapid Improvements in Reasoning","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2508.08636","citing_title":"InternBootcamp Technical Report: Boosting LLM Reasoning with Verifiable Task Scaling","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07731","citing_title":"Benchmarking EngGPT2-16B-A3B against Comparable Italian and International Open-source LLMs","ref_index":80,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19597","citing_title":"LLMEval-Logic: A Solver-Verified Chinese Benchmark for Logical Reasoning of LLMs with Adversarial Hardening","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15207","citing_title":"TeamTR: Trust-Region Fine-Tuning for Multi-Agent LLM Coordination","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2508.19035","citing_title":"Investigating Advanced Reasoning of Large Language Models via Black-Box Environment Interaction","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2505.24298","citing_title":"AReaL: A Large-Scale Asynchronous Reinforcement Learning System for Language Reasoning","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2511.20347","citing_title":"Soft Adaptive Policy Optimization","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2506.13585","citing_title":"MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08905","citing_title":"Forge: Quality-Aware Reinforcement Learning for NP-Hard Optimization in LLMs","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08498","citing_title":"MathConstraint: Automated Generation of Verified Combinatorial Reasoning Instances for LLMs","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05741","citing_title":"HyperLens: Quantifying Cognitive Effort in LLMs with Fine-grained Confidence Trajectory","ref_index":52,"is_internal_anchor":false},{"citing_arxiv_id":"2604.08477","citing_title":"SUPERNOVA: Eliciting General Reasoning in LLMs with Reinforcement Learning on Natural Instructions","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07731","citing_title":"Benchmarking EngGPT2-16B-A3B against Comparable Italian and International Open-source LLMs","ref_index":80,"is_internal_anchor":false},{"citing_arxiv_id":"2509.17765","citing_title":"Qwen3-Omni Technical Report","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2507.20534","citing_title":"Kimi K2: Open Agentic Intelligence","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15602","citing_title":"GroupDPO: Memory efficient Group-wise Direct Preference Optimization","ref_index":20,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE","json":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE.json","graph_json":"https://pith.science/api/pith-number/DFVYWZHKCJEWOUNW3PRJFZK2CE/graph.json","events_json":"https://pith.science/api/pith-number/DFVYWZHKCJEWOUNW3PRJFZK2CE/events.json","paper":"https://pith.science/paper/DFVYWZHK"},"agent_actions":{"view_html":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE","download_json":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE.json","view_paper":"https://pith.science/paper/DFVYWZHK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.01100&json=true","fetch_graph":"https://pith.science/api/pith-number/DFVYWZHKCJEWOUNW3PRJFZK2CE/graph.json","fetch_events":"https://pith.science/api/pith-number/DFVYWZHKCJEWOUNW3PRJFZK2CE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE/action/storage_attestation","attest_author":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE/action/author_attestation","sign_citation":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE/action/citation_signature","submit_replication":"https://pith.science/pith/DFVYWZHKCJEWOUNW3PRJFZK2CE/action/replication_record"}},"created_at":"2026-07-05T11:37:15.554405+00:00","updated_at":"2026-07-05T11:37:15.554405+00:00"}