{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:MKGPHTNO7EBXTIHQROOHMD52KB","short_pith_number":"pith:MKGPHTNO","schema_version":"1.0","canonical_sha256":"628cf3cdaef90379a0f08b9c760fba505f7e8e04684726cdef8ea547872a5fa5","source":{"kind":"arxiv","id":"1806.02847","version":2},"attestation_state":"computed","paper":{"title":"A Simple Method for Commonsense Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Quoc V. Le, Trieu H. Trinh","submitted_at":"2018-06-07T18:13:08Z","abstract_excerpt":"Commonsense reasoning is a long-standing challenge for deep learning. For example, it is difficult to use neural networks to tackle the Winograd Schema dataset (Levesque et al., 2011). In this paper, we present a simple method for commonsense reasoning with neural networks, using unsupervised learning. Key to our method is the use of language models, trained on a massive amount of unlabled data, to score multiple choice questions posed by commonsense reasoning tests. On both Pronoun Disambiguation and Winograd Schema challenges, our models outperform previous state-of-the-art methods by a larg"},"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":"1806.02847","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2018-06-07T18:13:08Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"5cba51a6d7af1d5bdad5d4e33053301cd1e93b66c25f748efae3c2d6119aabf5","abstract_canon_sha256":"e692fb9639ee9cecd206a3bebf22b984d31073b508c22a99b35158ccb9dcec46"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:07:34.298882Z","signature_b64":"44RQvmRaDV1XTwT2uM9pZWSX0T2rixmVq42jzkusu/q18GadhmK1ixs4ubxBd4ZEJYoNjZQB9bbEk/H8tFCyAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"628cf3cdaef90379a0f08b9c760fba505f7e8e04684726cdef8ea547872a5fa5","last_reissued_at":"2026-07-05T00:07:34.298404Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:07:34.298404Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Simple Method for Commonsense Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Quoc V. Le, Trieu H. Trinh","submitted_at":"2018-06-07T18:13:08Z","abstract_excerpt":"Commonsense reasoning is a long-standing challenge for deep learning. For example, it is difficult to use neural networks to tackle the Winograd Schema dataset (Levesque et al., 2011). In this paper, we present a simple method for commonsense reasoning with neural networks, using unsupervised learning. Key to our method is the use of language models, trained on a massive amount of unlabled data, to score multiple choice questions posed by commonsense reasoning tests. On both Pronoun Disambiguation and Winograd Schema challenges, our models outperform previous state-of-the-art methods by a larg"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1806.02847","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/1806.02847/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":"1806.02847","created_at":"2026-07-05T00:07:34.298463+00:00"},{"alias_kind":"arxiv_version","alias_value":"1806.02847v2","created_at":"2026-07-05T00:07:34.298463+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1806.02847","created_at":"2026-07-05T00:07:34.298463+00:00"},{"alias_kind":"pith_short_12","alias_value":"MKGPHTNO7EBX","created_at":"2026-07-05T00:07:34.298463+00:00"},{"alias_kind":"pith_short_16","alias_value":"MKGPHTNO7EBXTIHQ","created_at":"2026-07-05T00:07:34.298463+00:00"},{"alias_kind":"pith_short_8","alias_value":"MKGPHTNO","created_at":"2026-07-05T00:07:34.298463+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":20,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.28551","citing_title":"DataComp-VLM: Improved Open Datasets for Vision-Language Models","ref_index":283,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28551","citing_title":"DataComp-VLM: Improved Open Datasets for Vision-Language Models","ref_index":283,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30775","citing_title":"A Single Rewrite Suffices: Empirical Lessons from Production Skill Description Optimization","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2107.06499","citing_title":"Deduplicating Training Data Makes Language Models Better","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2201.11990","citing_title":"Using DeepSpeed and Megatron to Train Megatron-Turing NLG 530B, A Large-Scale Generative Language Model","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"1909.05858","citing_title":"CTRL: A Conditional Transformer Language Model for Controllable Generation","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2311.16867","citing_title":"The Falcon Series of Open Language Models","ref_index":198,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14360","citing_title":"A Formative Study of Brief Affective Text as a Complement to Wearable Sensing for Longitudinal Student Health Monitoring","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2306.01116","citing_title":"The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Only","ref_index":40,"is_internal_anchor":false},{"citing_arxiv_id":"2208.07339","citing_title":"LLM.int8(): 8-bit Matrix Multiplication for Transformers at Scale","ref_index":164,"is_internal_anchor":false},{"citing_arxiv_id":"2006.03654","citing_title":"DeBERTa: Decoding-enhanced BERT with Disentangled Attention","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"1910.10683","citing_title":"Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer","ref_index":72,"is_internal_anchor":false},{"citing_arxiv_id":"2206.07682","citing_title":"Emergent Abilities of Large Language Models","ref_index":87,"is_internal_anchor":false},{"citing_arxiv_id":"2303.18223","citing_title":"A Survey of Large Language Models","ref_index":162,"is_internal_anchor":false},{"citing_arxiv_id":"2101.00027","citing_title":"The Pile: An 800GB Dataset of Diverse Text for Language Modeling","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2205.01068","citing_title":"OPT: Open Pre-trained Transformer Language Models","ref_index":285,"is_internal_anchor":false},{"citing_arxiv_id":"1909.08053","citing_title":"Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2004.05150","citing_title":"Longformer: The Long-Document Transformer","ref_index":120,"is_internal_anchor":false},{"citing_arxiv_id":"2005.14165","citing_title":"Language Models are Few-Shot Learners","ref_index":78,"is_internal_anchor":false},{"citing_arxiv_id":"1907.11692","citing_title":"RoBERTa: A Robustly Optimized BERT Pretraining Approach","ref_index":42,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB","json":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB.json","graph_json":"https://pith.science/api/pith-number/MKGPHTNO7EBXTIHQROOHMD52KB/graph.json","events_json":"https://pith.science/api/pith-number/MKGPHTNO7EBXTIHQROOHMD52KB/events.json","paper":"https://pith.science/paper/MKGPHTNO"},"agent_actions":{"view_html":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB","download_json":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB.json","view_paper":"https://pith.science/paper/MKGPHTNO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1806.02847&json=true","fetch_graph":"https://pith.science/api/pith-number/MKGPHTNO7EBXTIHQROOHMD52KB/graph.json","fetch_events":"https://pith.science/api/pith-number/MKGPHTNO7EBXTIHQROOHMD52KB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB/action/storage_attestation","attest_author":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB/action/author_attestation","sign_citation":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB/action/citation_signature","submit_replication":"https://pith.science/pith/MKGPHTNO7EBXTIHQROOHMD52KB/action/replication_record"}},"created_at":"2026-07-05T00:07:34.298463+00:00","updated_at":"2026-07-05T00:07:34.298463+00:00"}