{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:FFT6PXEU6JVHN3VRTSPH7DYODG","short_pith_number":"pith:FFT6PXEU","schema_version":"1.0","canonical_sha256":"2967e7dc94f26a76eeb19c9e7f8f0e1992a98e7342e81fe9c0dfed4a5b237c27","source":{"kind":"arxiv","id":"1911.00172","version":2},"attestation_state":"computed","paper":{"title":"Generalization through Memorization: Nearest Neighbor Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Jurafsky, Luke Zettlemoyer, Mike Lewis, Omer Levy, Urvashi Khandelwal","submitted_at":"2019-11-01T01:09:53Z","abstract_excerpt":"We introduce $k$NN-LMs, which extend a pre-trained neural language model (LM) by linearly interpolating it with a $k$-nearest neighbors ($k$NN) model. The nearest neighbors are computed according to distance in the pre-trained LM embedding space, and can be drawn from any text collection, including the original LM training data. Applying this augmentation to a strong Wikitext-103 LM, with neighbors drawn from the original training set, our $k$NN-LM achieves a new state-of-the-art perplexity of 15.79 - a 2.9 point improvement with no additional training. We also show that this approach has impl"},"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":"1911.00172","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-11-01T01:09:53Z","cross_cats_sorted":[],"title_canon_sha256":"12761e99e360fb75e7d4928602e47fe7cd4a5d20f62fcbad4d821308cfa4dd3c","abstract_canon_sha256":"73531ebc23e73d63ee031c4275d967450b5067f2b67e8cc4245554b8c5708183"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:41:00.167761Z","signature_b64":"E0iRzwUyTeqnzGdOwFz+SRXolZ3CxP2BKDqlaZjFYL3Seow/YAqyb5ZjfxAVQFTlb5dJQTdosi6/4n5q/ZyPAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2967e7dc94f26a76eeb19c9e7f8f0e1992a98e7342e81fe9c0dfed4a5b237c27","last_reissued_at":"2026-07-05T00:41:00.167330Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:41:00.167330Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Generalization through Memorization: Nearest Neighbor Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Dan Jurafsky, Luke Zettlemoyer, Mike Lewis, Omer Levy, Urvashi Khandelwal","submitted_at":"2019-11-01T01:09:53Z","abstract_excerpt":"We introduce $k$NN-LMs, which extend a pre-trained neural language model (LM) by linearly interpolating it with a $k$-nearest neighbors ($k$NN) model. The nearest neighbors are computed according to distance in the pre-trained LM embedding space, and can be drawn from any text collection, including the original LM training data. Applying this augmentation to a strong Wikitext-103 LM, with neighbors drawn from the original training set, our $k$NN-LM achieves a new state-of-the-art perplexity of 15.79 - a 2.9 point improvement with no additional training. We also show that this approach has impl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1911.00172","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/1911.00172/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":"1911.00172","created_at":"2026-07-05T00:41:00.167390+00:00"},{"alias_kind":"arxiv_version","alias_value":"1911.00172v2","created_at":"2026-07-05T00:41:00.167390+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1911.00172","created_at":"2026-07-05T00:41:00.167390+00:00"},{"alias_kind":"pith_short_12","alias_value":"FFT6PXEU6JVH","created_at":"2026-07-05T00:41:00.167390+00:00"},{"alias_kind":"pith_short_16","alias_value":"FFT6PXEU6JVHN3VR","created_at":"2026-07-05T00:41:00.167390+00:00"},{"alias_kind":"pith_short_8","alias_value":"FFT6PXEU","created_at":"2026-07-05T00:41:00.167390+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":30,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.17162","citing_title":"MemSlides: A Hierarchical Memory Driven Agent Framework for Personalized Slide Generation with Multi-turn Local Revision","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2606.13001","citing_title":"CFALR: Collaborative Filtering-Augmented Large Language Model for Personalized Fashion Outfit Recommendation","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02010","citing_title":"InduceKV: Fixed-Footprint Continual Adaptation of Multimodal LLMs via Inducing KV Memories","ref_index":204,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02303","citing_title":"A Hippocampus for Linear Attention: An Exact Memory for What the Recurrent State Forgets","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2606.05936","citing_title":"Epistemic Injustice in Language Models: An Audit of Pretraining Filters and Guardrails","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2606.28876","citing_title":"Memory-Managed Long-Context Attention: Bounded Editable Memory with a Hard Lifecycle and Calibrated Sparse Fallback","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25716","citing_title":"An Efficient and Privacy-Preserving Architecture for Cross-Institutional Collaborative RAG","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27686","citing_title":"Tensor Memory: Fixed-Size Recurrent State for Long-Horizon Transformers","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17630","citing_title":"SegRAG: Training-Free Retrieval-Augmented Semantic Segmentation","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08143","citing_title":"HoReN: Normalized Hopfield Retrieval for Large-Scale Sequential Model Editing","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17630","citing_title":"SegRAG: Training-Free Retrieval-Augmented Semantic Segmentation","ref_index":57,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16893","citing_title":"NGM: A Plug-and-Play Training-Free Memory Module for LLMs","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2510.17934","citing_title":"AtlasKV: Augmenting LLMs with Billion-Scale Knowledge Graphs in 20GB VRAM","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2510.26083","citing_title":"Nirvana: A Specialized Generalist Model With Task-Aware Memory Mechanism","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2602.11183","citing_title":"Mitigating Error Accumulation in Continuous Navigation via Memory-Augmented Kalman Filtering","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2309.16671","citing_title":"Demystifying CLIP Data","ref_index":98,"is_internal_anchor":false},{"citing_arxiv_id":"2401.18059","citing_title":"RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval","ref_index":150,"is_internal_anchor":false},{"citing_arxiv_id":"2603.10126","citing_title":"AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2002.08909","citing_title":"REALM: Retrieval-Augmented Language Model Pre-Training","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2303.16199","citing_title":"LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention","ref_index":267,"is_internal_anchor":false},{"citing_arxiv_id":"2201.08239","citing_title":"LaMDA: Language Models for Dialog Applications","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06225","citing_title":"Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06216","citing_title":"TIDE: Every Layer Knows the Token Beneath the Context","ref_index":99,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06225","citing_title":"Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04651","citing_title":"FAAST: Forward-Only Associative Learning via Closed-Form Fast Weights for Test-Time Supervised Adaptation","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG","json":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG.json","graph_json":"https://pith.science/api/pith-number/FFT6PXEU6JVHN3VRTSPH7DYODG/graph.json","events_json":"https://pith.science/api/pith-number/FFT6PXEU6JVHN3VRTSPH7DYODG/events.json","paper":"https://pith.science/paper/FFT6PXEU"},"agent_actions":{"view_html":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG","download_json":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG.json","view_paper":"https://pith.science/paper/FFT6PXEU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1911.00172&json=true","fetch_graph":"https://pith.science/api/pith-number/FFT6PXEU6JVHN3VRTSPH7DYODG/graph.json","fetch_events":"https://pith.science/api/pith-number/FFT6PXEU6JVHN3VRTSPH7DYODG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG/action/storage_attestation","attest_author":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG/action/author_attestation","sign_citation":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG/action/citation_signature","submit_replication":"https://pith.science/pith/FFT6PXEU6JVHN3VRTSPH7DYODG/action/replication_record"}},"created_at":"2026-07-05T00:41:00.167390+00:00","updated_at":"2026-07-05T00:41:00.167390+00:00"}