{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:MTDDT7RZZZSNFR42RLMH64RFHQ","short_pith_number":"pith:MTDDT7RZ","schema_version":"1.0","canonical_sha256":"64c639fe39ce64d2c79a8ad87f72253c0287309588b0bf07b4474194e38c5d24","source":{"kind":"arxiv","id":"2501.03035","version":4},"attestation_state":"computed","paper":{"title":"Quantization Meets Reasoning: Exploring LLM Low-Bit Quantization Degradation for Mathematical Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Congkai Xie, Hongxia Yang, Ngai Wong, Runming Yang, Yupeng Su, Zheng Wang, Zhen Li, Zhongwei Xie","submitted_at":"2025-01-06T14:23:02Z","abstract_excerpt":"Large language models have achieved significant advancements in complex mathematical reasoning benchmarks, such as MATH. However, their substantial computational requirements present challenges for practical deployment. Model quantization has emerged as an effective strategy to reduce memory usage and computational costs by employing lower precision and bit-width representations. In this study, we systematically evaluate the impact of quantization on mathematical reasoning tasks. Our results demonstrate that aggressive quantization methods like AWQ and GPTQ introduce up to 32.39% accuracy degr"},"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":"2501.03035","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-01-06T14:23:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"af938ed86069c410f0f03a2bf798a8e7f3859600af5b65f08a6a095b76101f59","abstract_canon_sha256":"f150da0a9290ea4871576b2051f88fb6c86ce89a4d2618cce2df05f38dc4ecbc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:18:47.930412Z","signature_b64":"0zQo+BPXcEGcW9r+TUkfIpdK/AgUlyFN+vRsbjHtD/UfXbvyG3yTg5fx1B5toKGnsZ72/jUXJxEDoT6Qdy7jCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"64c639fe39ce64d2c79a8ad87f72253c0287309588b0bf07b4474194e38c5d24","last_reissued_at":"2026-07-05T10:18:47.929904Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:18:47.929904Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Quantization Meets Reasoning: Exploring LLM Low-Bit Quantization Degradation for Mathematical Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Congkai Xie, Hongxia Yang, Ngai Wong, Runming Yang, Yupeng Su, Zheng Wang, Zhen Li, Zhongwei Xie","submitted_at":"2025-01-06T14:23:02Z","abstract_excerpt":"Large language models have achieved significant advancements in complex mathematical reasoning benchmarks, such as MATH. However, their substantial computational requirements present challenges for practical deployment. Model quantization has emerged as an effective strategy to reduce memory usage and computational costs by employing lower precision and bit-width representations. In this study, we systematically evaluate the impact of quantization on mathematical reasoning tasks. Our results demonstrate that aggressive quantization methods like AWQ and GPTQ introduce up to 32.39% accuracy degr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.03035","kind":"arxiv","version":4},"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/2501.03035/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":"2501.03035","created_at":"2026-07-05T10:18:47.929967+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.03035v4","created_at":"2026-07-05T10:18:47.929967+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.03035","created_at":"2026-07-05T10:18:47.929967+00:00"},{"alias_kind":"pith_short_12","alias_value":"MTDDT7RZZZSN","created_at":"2026-07-05T10:18:47.929967+00:00"},{"alias_kind":"pith_short_16","alias_value":"MTDDT7RZZZSNFR42","created_at":"2026-07-05T10:18:47.929967+00:00"},{"alias_kind":"pith_short_8","alias_value":"MTDDT7RZ","created_at":"2026-07-05T10:18:47.929967+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2607.08734","citing_title":"The Illusion of Equivalency: Statistical Characterization of Quantization Effects in LLMs","ref_index":30,"is_internal_anchor":true},{"citing_arxiv_id":"2607.08734","citing_title":"The Illusion of Equivalency: Statistical Characterization of Quantization Effects in LLMs","ref_index":73,"is_internal_anchor":true},{"citing_arxiv_id":"2606.25674","citing_title":"BitNet Text Embeddings","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00908","citing_title":"Beyond Activation Alignment:The Alignment-Diversity Tradeoff in Task-Aware LLM Quantization","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00206","citing_title":"Quantized Reasoning Models Think They Need to Think Longer, but They Do Not","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20315","citing_title":"Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2509.07177","citing_title":"Towards EnergyGPT: A Large Language Model Specialized for the Energy Sector","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2602.13595","citing_title":"The Quantization Trap: Breaking Linear Scaling Laws in Multi-Hop Reasoning","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.02715","citing_title":"FluxMoE: Decoupling Expert Residency for High-Performance MoE Serving","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08568","citing_title":"Different Prompts, Different Ranks: Prompt-aware Dynamic Rank Selection for SVD-based LLM Compression","ref_index":61,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08755","citing_title":"LAQuant: A Simple Overhead-free Large Reasoning Model Quantization by Layer-wise Lookahead Loss","ref_index":29,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ","json":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ.json","graph_json":"https://pith.science/api/pith-number/MTDDT7RZZZSNFR42RLMH64RFHQ/graph.json","events_json":"https://pith.science/api/pith-number/MTDDT7RZZZSNFR42RLMH64RFHQ/events.json","paper":"https://pith.science/paper/MTDDT7RZ"},"agent_actions":{"view_html":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ","download_json":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ.json","view_paper":"https://pith.science/paper/MTDDT7RZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.03035&json=true","fetch_graph":"https://pith.science/api/pith-number/MTDDT7RZZZSNFR42RLMH64RFHQ/graph.json","fetch_events":"https://pith.science/api/pith-number/MTDDT7RZZZSNFR42RLMH64RFHQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ/action/storage_attestation","attest_author":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ/action/author_attestation","sign_citation":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ/action/citation_signature","submit_replication":"https://pith.science/pith/MTDDT7RZZZSNFR42RLMH64RFHQ/action/replication_record"}},"created_at":"2026-07-05T10:18:47.929967+00:00","updated_at":"2026-07-05T10:18:47.929967+00:00"}