REVIEW 38 cited by
Generalization through Memorization: Nearest Neighbor Language Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Generalization through Memorization: Nearest Neighbor Language Models
read the original abstract
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 implications for efficiently scaling up to larger training sets and allows for effective domain adaptation, by simply varying the nearest neighbor datastore, again without further training. Qualitatively, the model is particularly helpful in predicting rare patterns, such as factual knowledge. Together, these results strongly suggest that learning similarity between sequences of text is easier than predicting the next word, and that nearest neighbor search is an effective approach for language modeling in the long tail.
Forward citations
Cited by 38 Pith papers
-
REALM: Retrieval-Augmented Language Model Pre-Training
REALM augments language-model pre-training with an unsupervised retriever over Wikipedia documents and reports 4-16% absolute gains on open-domain QA benchmarks over prior implicit and explicit knowledge methods.
-
Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs
Memory Inception steers LLMs via selective latent KV cache injection at chosen layers, delivering better control-drift balance than prompting or CAA on personality and reasoning tasks while reducing storage needs.
-
AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models
AR-VLA introduces a standalone autoregressive action expert with long-lived memory that generates context-aware continuous actions for VLAs, replacing chunk-based heads with smoother trajectories and maintained task success.
-
Mitigating Error Accumulation in Continuous Navigation via Memory-Augmented Kalman Filtering
NeuroKalman mitigates state drift in vision-language UAV navigation by using memory-augmented Kalman filtering where attention retrieves historical anchors to correct predictions without gradient updates.
-
LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention
LLaMA-Adapter turns frozen LLaMA 7B into a capable instruction follower using only 1.2M new parameters and zero-init attention, matching Alpaca while extending to image-conditioned reasoning on ScienceQA and COCO.
-
Megatron-LM: Training Multi-Billion Parameter Language Models Using Model Parallelism
Intra-layer model parallelism in PyTorch enables training of 8.3B-parameter transformers, achieving SOTA perplexity of 10.8 on WikiText103 and 66.5% accuracy on LAMBADA.
-
MemSFT: Mitigating Alignment Tax with an External Parametric Memory
MemSFT attaches a retriever-imitating 8B memory plus a word-level router to frozen Qwen3 backbones, boosting domain scores by ~36 points while holding general-benchmark averages essentially flat, where full SFT loses ...
-
Remembering Distinct Items, Not Tokens: A Learnable Dirichlet-Process Cache Between State-Space Models and Attention
A DP-means allocate-on-novelty cache matches full-attention associative recall while storing only distinct items, and a minimal novelty gate recovers the rule end-to-end.
-
Exploiting Structural Properties for Efficient Constraint-Aware HNSW Hyperparameter Tuning
CHAT uses HNSW-specific monotonic and unimodal structure plus resource surrogates to tune M, efc, and efs under constraints, beating black-box tuners by up to 45% throughput or 11% recall and up to 44× faster convergence.
-
Text as Partial Constraint: Core-Residual Alignment for Robust Vision-Language Learning
Aligning images to multi-view caption cores while suppressing orthogonal residual text and disagreement-aware temperature improves robust zero-shot recognition and LVLM transfer.
-
A Hippocampus for Linear Attention: An Exact Memory for What the Recurrent State Forgets
HOLA pairs a compressive delta-rule recurrent state with a residual-selected exact KV cache and decoupled RMSNorm-gamma read, yielding lower perplexity than both standard linear attention and full-attention baselines ...
-
InduceKV: Fixed-Footprint Continual Adaptation of Multimodal LLMs via Inducing KV Memories
InduceKV is a retrieval-based continual adaptation method that uses bilevel selection to build a compact set of inducing KV memories for fixed-footprint updates to multimodal LLMs.
-
Memory-Managed Long-Context Attention: Bounded Editable Memory with a Hard Lifecycle and Calibrated Sparse Fallback
A hybrid attention mechanism with editable request-local memory slots and sparse fallback achieves high accuracy on synthetic overwrite, version, and anti-pollution tasks where pure fixed-state or sparse methods fail,...
-
Tensor Memory: Fixed-Size Recurrent State for Long-Horizon Transformers
Tensor Memory augments Transformers with a constant-size 3D voxel grid using differentiable soft writes at predicted locations, local interaction, and gated recurrent dynamics to decouple memory capacity from sequence length.
-
An Efficient and Privacy-Preserving Architecture for Cross-Institutional Collaborative RAG
FedRAG uses a Scrambled Distributed Attention protocol with feature scrambling and token permutation to enable high-throughput, privacy-preserving federated RAG without special hardware or retraining.
-
SegRAG: Training-Free Retrieval-Augmented Semantic Segmentation
SegRAG is a training-free retrieval-augmented framework that extracts class-specific point prompts from a filtered DINOv3 feature bank to boost SAM3 semantic segmentation performance on standard and agricultural benchmarks.
-
SegRAG: Training-Free Retrieval-Augmented Semantic Segmentation
SegRAG augments SAM3 with class-specific point prompts retrieved via DINOv3 features and filtered by ICCD, using TSG at inference to improve open-vocabulary segmentation.
-
Memory Inception: Latent-Space KV Cache Manipulation for Steering LLMs
Memory Inception is a training-free method that injects latent KV banks at chosen layers to steer LLMs, achieving superior control-drift balance and up to 118x storage reduction on personality and structured-reasoning tasks.
-
FAAST: Forward-Only Associative Learning via Closed-Form Fast Weights for Test-Time Supervised Adaptation
FAAST analytically compiles labeled examples into fast weights via a single forward pass, matching backprop adaptation performance with over 90% less time and up to 95% less memory than memory-based methods.
-
HoReN: Normalized Hopfield Retrieval for Large-Scale Sequential Model Editing
HoReN achieves stable sequential editing of 50K facts in LLMs by combining a normalized Hopfield codebook with angular retrieval and attractor dynamics.
-
Adaptive Defense Orchestration for RAG: A Sentinel-Strategist Architecture against Multi-Vector Attacks
A context-aware Sentinel-Strategist system for RAG selectively applies defenses to block membership inference and data poisoning while recovering most retrieval utility compared to always-on defense stacks.
-
Cram Less to Fit More: Training Data Pruning Improves Memorization of Facts
Loss-based pruning of training data to limit facts and flatten their frequency distribution enables a 110M-parameter GPT-2 model to memorize 1.3 times more entity facts than standard training, matching a 1.3B-paramete...
-
AtlasKV: Augmenting LLMs with Billion-Scale Knowledge Graphs in 20GB VRAM
AtlasKV integrates billion-scale KGs into LLMs parametrically with sub-linear complexity and low memory by converting triples into key-value representations handled by the model's attention.
-
RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval
RAPTOR introduces a tree-organized retrieval method using recursive abstractive summaries, achieving a 20% absolute accuracy improvement on the QuALITY benchmark when paired with GPT-4.
-
Demystifying CLIP Data
MetaCLIP curates balanced 400M-pair subsets from CommonCrawl that outperform CLIP data, reaching 70.8% zero-shot ImageNet accuracy on ViT-B versus CLIP's 68.3%.
-
LaMDA: Language Models for Dialog Applications
LaMDA shows that fine-tuning on human-value annotations and consulting external knowledge sources significantly improves safety and factual grounding in large dialog models beyond what scaling alone achieves.
-
Memory Decoder at Scale: A Pretrained, Parametric Long-Term Memory
A pretrained parametric memory module, taught to copy nearest-neighbor retrieval for next-token prediction, lets small frozen LMs match or beat much larger LMs at the same total parameter budget.
-
Memory-Managed Long-Context Attention: Bounded Editable Memory with a Hard Lifecycle and Calibrated Sparse Fallback
Editable bounded memory slots with lifecycle control plus sparse fallback cover overwrite and no-signal long-context cases that pure fixed-state or pure sparse methods fail under controlled conditions.
-
MemSlides: A Hierarchical Memory Driven Agent Framework for Personalized Slide Generation with Multi-turn Local Revision
MemSlides introduces a three-part memory hierarchy (user profile, working, tool) with scoped local revision for multi-turn personalized slide generation.
-
CFALR: Collaborative Filtering-Augmented Large Language Model for Personalized Fashion Outfit Recommendation
CFALR augments LLMs with collaborative filtering embeddings via trainable projection layers to outperform prior CF and LLM methods on Polyvore and IQON for personalized outfit tasks.
-
Epistemic Injustice in Language Models: An Audit of Pretraining Filters and Guardrails
An audit finds language model filters and guardrails disproportionately suppress mentions of marginalized groups via lexical cues while failing to catch explicit harms.
-
NGM: A Plug-and-Play Training-Free Memory Module for LLMs
NGM is a plug-and-play n-gram memory module that encodes n-grams from pretrained embeddings and gates their injection to improve LLM performance by 0.5-1.2 points on average across eight benchmarks.
-
TIDE: Every Layer Knows the Token Beneath the Context
TIDE augments standard transformers with per-layer token embedding injection via an ensemble of memory blocks and a depth-conditioned router to mitigate rare-token undertraining and contextual collapse.
-
FAAST: Forward-Only Associative Learning via Closed-Form Fast Weights for Test-Time Supervised Adaptation
FAAST performs test-time supervised adaptation by analytically deriving fast weights from examples in one forward pass, matching backprop performance with over 90% less adaptation time and up to 95% memory savings ver...
-
HoReN: Normalized Hopfield Retrieval for Large-Scale Sequential Model Editing
HoReN is a parameter-preserving editor that wraps an MLP with a Hopfield codebook memory and scales to 50K sequential edits on ZsRE while maintaining performance above 0.93.
-
CogniConsole: Externalizing Inference-Time Control as a Formal Abstraction for Reliable LLM Interactions
Under a fixed model, increasing structural scaffolding of inference-time control systematically reduces LLM output variance and failure rates on controllability probes.
-
Nirvana: A Specialized Generalist Model With Task-Aware Memory Mechanism
Nirvana adds a task-aware memory trigger and updater to specialized generalist models, achieving strong general benchmark results, lowest perplexity in biomedicine/finance/law, and improved MRI reconstruction fidelity.
-
Enhancing Large Language Models with Retrieval Augmented Generation for Software Testing and Inspection Automation
RAG-enhanced LLMs show generally positive effects on automated test generation and code inspection by supplying supplementary context that reduces hallucinations.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.