{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:5N2N3GXHAVGJ3DIAERBA2ED736","short_pith_number":"pith:5N2N3GXH","schema_version":"1.0","canonical_sha256":"eb74dd9ae7054c9d8d0024420d107fdf956cf16e993c09c33a7d78713303065d","source":{"kind":"arxiv","id":"1905.07830","version":1},"attestation_state":"computed","paper":{"title":"HellaSwag: Can a Machine Really Finish Your Sentence?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"HellaSwag shows state-of-the-art models still fail at commonsense sentence completion that humans solve easily.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ali Farhadi, Ari Holtzman, Rowan Zellers, Yejin Choi, Yonatan Bisk","submitted_at":"2019-05-19T23:57:23Z","abstract_excerpt":"Recent work by Zellers et al. (2018) introduced a new task of commonsense natural language inference: given an event description such as \"A woman sits at a piano,\" a machine must select the most likely followup: \"She sets her fingers on the keys.\" With the introduction of BERT, near human-level performance was reached. Does this mean that machines can perform human level commonsense inference?\n  In this paper, we show that commonsense inference still proves difficult for even state-of-the-art models, by presenting HellaSwag, a new challenge dataset. Though its questions are trivial for humans "},"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":true,"formal_links_present":false},"canonical_record":{"source":{"id":"1905.07830","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-05-19T23:57:23Z","cross_cats_sorted":[],"title_canon_sha256":"e9e53d4b2c8e24587b3aa33ee5bfedbfafdf21af4617107e422e70fe02959910","abstract_canon_sha256":"927c4889782be6b6bd5c205399e62f8e4869cac9638c87bfc36eb6830589af43"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-04T23:38:28.264304Z","signature_b64":"8LYRE/EXerMYSN8Ff4AGCJDuWRzSl+GTvNhKaTd4lUOfx5Lc1RhWTOAgblWdvXrPdx3cV18/IrKLUmr6Af0YAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eb74dd9ae7054c9d8d0024420d107fdf956cf16e993c09c33a7d78713303065d","last_reissued_at":"2026-07-04T23:38:28.263807Z","signature_status":"signed_v1","first_computed_at":"2026-07-04T23:38:28.263807Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"HellaSwag: Can a Machine Really Finish Your Sentence?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"HellaSwag shows state-of-the-art models still fail at commonsense sentence completion that humans solve easily.","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ali Farhadi, Ari Holtzman, Rowan Zellers, Yejin Choi, Yonatan Bisk","submitted_at":"2019-05-19T23:57:23Z","abstract_excerpt":"Recent work by Zellers et al. (2018) introduced a new task of commonsense natural language inference: given an event description such as \"A woman sits at a piano,\" a machine must select the most likely followup: \"She sets her fingers on the keys.\" With the introduction of BERT, near human-level performance was reached. Does this mean that machines can perform human level commonsense inference?\n  In this paper, we show that commonsense inference still proves difficult for even state-of-the-art models, by presenting HellaSwag, a new challenge dataset. Though its questions are trivial for humans "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Though its questions are trivial for humans (>95% accuracy), state-of-the-art models struggle (<48%).","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That adversarial filtering produces examples requiring genuine commonsense reasoning rather than merely exploiting specific weaknesses or distributional artifacts in the models used during filtering.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"HellaSwag dataset shows state-of-the-art models fail commonsense inference tasks that humans solve easily, built via adversarial filtering of distractors.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"HellaSwag shows state-of-the-art models still fail at commonsense sentence completion that humans solve easily.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"0aba65de350fc7c765ee8f342d88f717b5e1797142ba6879a16035483c3f1f3c"},"source":{"id":"1905.07830","kind":"arxiv","version":1},"verdict":{"id":"4928dd65-3589-403a-bee7-d0205863269c","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-11T02:51:33.151353Z","strongest_claim":"Though its questions are trivial for humans (>95% accuracy), state-of-the-art models struggle (<48%).","one_line_summary":"HellaSwag dataset shows state-of-the-art models fail commonsense inference tasks that humans solve easily, built via adversarial filtering of distractors.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That adversarial filtering produces examples requiring genuine commonsense reasoning rather than merely exploiting specific weaknesses or distributional artifacts in the models used during filtering.","pith_extraction_headline":"HellaSwag shows state-of-the-art models still fail at commonsense sentence completion that humans solve easily."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/1905.07830/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":19,"sample":[{"doi":"","year":2018,"title":"Yonatan Belinkov and Yonatan Bisk. 2018. Synthetic and natural noise both break neural machine translation. In ICLR. ICLR","work_id":"2f87ef9e-973a-4f33-93e7-a79e927ca32e","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2017,"title":"Qian Chen, Xiaodan Zhu, Zhen-Hua Ling, Si Wei, Hui Jiang, and Diana Inkpen. 2017. Enhanced lstm for natural language inference. In Proceedings of the 55th Annual Meeting of the Association for Computa","work_id":"8bd6683b-0e73-4ec0-bb86-391a37e265de","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding","work_id":"ed240a10-5b19-406c-baa5-30803f465785","ref_index":3,"cited_arxiv_id":"1810.04805","is_internal_anchor":true},{"doi":"","year":2018,"title":"Max Glockner, Vered Shwartz, and Yoav Goldberg. 2018. Breaking nli systems with sentences that require simple lexical inferences. In Proceedings of the 56th Annual Meeting of the Association for Compu","work_id":"277326a1-be96-4bb7-aab1-64021dbd9013","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2013,"title":"Jonathan Gordon and Benjamin Van Durme. 2013. Reporting bias and knowledge acquisition. In Proceedings of the 2013 workshop on Automated knowledge base construction, pages 25--30. ACM","work_id":"1002e502-b170-4aca-bbbd-64a081e2f89b","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":19,"snapshot_sha256":"fc7576631fb4f5e083bb1553be3225ef4ffe26fe3c0026780127e85c0555c0e6","internal_anchors":2},"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":"1905.07830","created_at":"2026-07-04T23:38:28.263867+00:00"},{"alias_kind":"arxiv_version","alias_value":"1905.07830v1","created_at":"2026-07-04T23:38:28.263867+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1905.07830","created_at":"2026-07-04T23:38:28.263867+00:00"},{"alias_kind":"pith_short_12","alias_value":"5N2N3GXHAVGJ","created_at":"2026-07-04T23:38:28.263867+00:00"},{"alias_kind":"pith_short_16","alias_value":"5N2N3GXHAVGJ3DIA","created_at":"2026-07-04T23:38:28.263867+00:00"},{"alias_kind":"pith_short_8","alias_value":"5N2N3GXH","created_at":"2026-07-04T23:38:28.263867+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":154,"internal_anchor_count":154,"sample":[{"citing_arxiv_id":"2606.25331","citing_title":"Improved Large Language Diffusion Models","ref_index":34,"is_internal_anchor":true},{"citing_arxiv_id":"2606.26587","citing_title":"SharQ: Bridging Activation Sparsity and FP4 Quantization for LLM Inference","ref_index":82,"is_internal_anchor":true},{"citing_arxiv_id":"2606.23419","citing_title":"GRINQH: Graded Input-based Quantization Hierarchy for Efficient LLM Generation","ref_index":47,"is_internal_anchor":true},{"citing_arxiv_id":"2606.23670","citing_title":"Tapered Language Models","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2606.22249","citing_title":"On the Expressive Power of Weight Quantization in Large Language Models","ref_index":45,"is_internal_anchor":true},{"citing_arxiv_id":"2606.22189","citing_title":"L20-Edu-135M: An Auditable Single-GPU Study of Data-Efficient Small Language Modeling","ref_index":22,"is_internal_anchor":true},{"citing_arxiv_id":"2606.21847","citing_title":"UniRank: Unified Rank Allocation for Low-Rank LLM Compression","ref_index":37,"is_internal_anchor":true},{"citing_arxiv_id":"2606.21228","citing_title":"Sakana Fugu Technical Report","ref_index":200,"is_internal_anchor":true},{"citing_arxiv_id":"2606.18650","citing_title":"BLADE: Scalable Bi-level Adaptive Data Selection for LLM Training","ref_index":51,"is_internal_anchor":true},{"citing_arxiv_id":"2606.18383","citing_title":"From Sparse Features to Trustworthy Proxies: Certifying SAE-Based Interpretability","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2607.01678","citing_title":"SCAPE: Accurate and Efficient LLM Training with Extreme Sparse Communication","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2607.02464","citing_title":"Will Scaling Improve Social Simulation with LLMs?","ref_index":43,"is_internal_anchor":true},{"citing_arxiv_id":"2606.13054","citing_title":"TWLA: Achieving Ternary Weights and Low-Bit Activations for LLMs via Post-Training Quantization","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2606.09080","citing_title":"Beyond FLOPs: Benchmarking Real Inference Acceleration of LLM Pruning under a GEMM-Centric Taxonomy","ref_index":64,"is_internal_anchor":true},{"citing_arxiv_id":"2606.08814","citing_title":"STAR: Rethinking MoE Routing as Structure-Aware Subspace Learning","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2606.08804","citing_title":"Q-Delta: Beyond Key-Value Associative State Evolution","ref_index":92,"is_internal_anchor":true},{"citing_arxiv_id":"2607.01065","citing_title":"GSRQ: Gain-Shape Residual Quantization for Sub-1-bit KV Cache","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2606.06547","citing_title":"FAIR-Calib: Frontier-Aware Instability-Reweighted Calibration for Post-Training Quantization of Diffusion Large Language Models","ref_index":19,"is_internal_anchor":true},{"citing_arxiv_id":"2606.05014","citing_title":"Depth-Attention: Cross-Layer Value Mixing for Language Models","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2606.04980","citing_title":"AlphaQ: Calibration-Free Bit Allocation for Mixture-of-Experts Quantization","ref_index":24,"is_internal_anchor":true},{"citing_arxiv_id":"2606.04050","citing_title":"LiftQuant: Continuous Bit-Width LLM via Dimensional Lifting and Projection","ref_index":32,"is_internal_anchor":true},{"citing_arxiv_id":"2606.02430","citing_title":"Not All Errors Are Equal: A Systematic Study of Error Propagation in Large Language Model Inference","ref_index":95,"is_internal_anchor":true},{"citing_arxiv_id":"2606.02780","citing_title":"Do Value Vectors in Deep Layers Need Context from the Residual Stream?","ref_index":109,"is_internal_anchor":true},{"citing_arxiv_id":"2606.01666","citing_title":"DOT-MoE: Differentiable Optimal Transport for MoEfication","ref_index":21,"is_internal_anchor":true},{"citing_arxiv_id":"2606.31397","citing_title":"Mixture-of-Control: State-Aware Fine-Tuning for Transformer-based Models","ref_index":92,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736","json":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736.json","graph_json":"https://pith.science/api/pith-number/5N2N3GXHAVGJ3DIAERBA2ED736/graph.json","events_json":"https://pith.science/api/pith-number/5N2N3GXHAVGJ3DIAERBA2ED736/events.json","paper":"https://pith.science/paper/5N2N3GXH"},"agent_actions":{"view_html":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736","download_json":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736.json","view_paper":"https://pith.science/paper/5N2N3GXH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1905.07830&json=true","fetch_graph":"https://pith.science/api/pith-number/5N2N3GXHAVGJ3DIAERBA2ED736/graph.json","fetch_events":"https://pith.science/api/pith-number/5N2N3GXHAVGJ3DIAERBA2ED736/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736/action/storage_attestation","attest_author":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736/action/author_attestation","sign_citation":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736/action/citation_signature","submit_replication":"https://pith.science/pith/5N2N3GXHAVGJ3DIAERBA2ED736/action/replication_record"}},"created_at":"2026-07-04T23:38:28.263867+00:00","updated_at":"2026-07-04T23:38:28.263867+00:00"}