Proposes a method to train neural networks directly on compressed text data.
problem Training neural networks on compressed text data without decompression.
method Introduces composer modules to encode symbols from grammar compression rules into vector representations.
result Demonstrates that the proposed method can achieve both memory and computational efficiency while maintaining moderate performance.
Neural model extracts tokens as latent variables for text compression.
problem Compressing text using neural models.
method Extracting tokens with highest tf-idf scores or highest loss from a bidirectional language model.
result Extracting tokens as latent variables significantly outperforms state-of-the-art methods.
Can textual data be compressed intelligently without losing accuracy in evaluating sentiment? In this study, we propose a novel evolutionary compression algorithm, PARSEC (PARts-of-Speech for sEntiment Compression), which makes use of Parts-of-Speech tags to compress text in a way that sacrifices minimal classification…
Neural networks compress uninformative input directions, improving test error.
problem Data lie in a high-dimensional space but labels vary along a lower-dimensional manifold.
method One-hidden layer network trained with gradient descent, analyzing weight evolution and compression.
result Compression factor λ ∼ √p improves test error, with β Feature > β Lazy.
Compressed LLM embeddings improve noisy regression tasks without overfitting.
problem Noisy regression tasks with high signal-to-noise ratios.
method Comparison of embedding compression techniques using autoencoder hidden representations.
result Compression improves performance on noisy tasks like financial return prediction.
Improved image compression with diffusion models outperforming state-of-the-art methods.
problem Difficulties in replicating text-to-image success in image compression.
method Two-stage approach combining autoencoder targeting MSE followed by score-based decoder.
result Significantly improved perceptual quality at a given bit-rate, measured by FID score.
New bounds for large language models using token properties.
problem Vacuous generalization bounds for large language models.
method Martingale properties and Monarch matrices.
result Non-vacuous generalization bounds for LLMs up to 70B parameters.
This paper introduces a new Lagrangian for the Information Bottleneck problem to simplify optimization.
problem Optimizing compressed representations for predicting Y Y Y while limiting information about X X X . method Introduces a general family of Lagrangians to explore the Information Bottleneck curve.
result Solves the original constrained optimization problem with a single optimization.
Extractive compression is a challenging natural language processing problem. This work contributes by formulating neural extractive compression as a parse tree transduction problem, rather than a sequence transduction task. Motivated by this, we introduce a deep neural model for learning structure-to-substructure tree …
Differentiable mask prunes deep networks for vision and text.
problem Efficiently compressing deep networks for edge devices.
method Introduces a differentiable mask for sparsity induction.
result Successfully prunes weights, filters, and nodes of convolutional and recurrent networks.
This paper proposes to perform authorship analysis using the Fast Compression Distance (FCD), a similarity measure based on compression with dictionaries directly extracted from the written texts. The FCD computes a similarity between two documents through an effective binary search on the intersection set between the …
Data compression is a popular technique for improving the efficiency of data processing workloads such as SQL queries and more recently, machine learning (ML) with classical batch gradient methods. But the efficacy of such ideas for mini-batch stochastic gradient descent (MGD), arguably the workhorse algorithm of moder…
LLMs compress financial texts, but distort decision-making.
problem LLMs compress financial texts, altering decision-making.
method Analyzed two diagnostic patterns: decontextualization and model dependency. Proposed Agentic Context Compression.
result LLM-compressed financial texts alter decision-making.
Compression method reduces word embedding size for NLP models.
problem Memory constraints in deploying deep learning models for NLP tasks.
method Low rank matrix factorization during training to compress word embeddings.
result 90% compression with minimal accuracy loss for sentence classification tasks.
We discuss algorithms for estimating the Shannon entropy h of finite symbol sequences with long range correlations. In particular, we consider algorithms which estimate h from the code lengths produced by some compression algorithm. Our interest is in describing their convergence with sequence length, assuming no limit…
Discrete Flow Maps bypass sequential prediction limits for parallel text generation.
problem Sequential autoregressive prediction limits large language model speed.
method Flow Maps compress generative trajectories into single-step mappings.
result Discrete Flow Maps surpass previous state-of-the-art results in discrete flow modeling.
New compression methods handle biased input sequences for more accurate posterior summaries.
problem Handling biased input sequences for accurate posterior summaries.
method Stein kernel thinning, low-rank SKT, Stein recombination, Stein Cholesky.
result Achieves accurate posterior summaries with biased input sequences.
Weight Squeezing transfers knowledge from large models to smaller ones, improving performance and speed.
problem Transfer learning and model compression for faster and more efficient training.
method Reparameterization of weights from a large model to a smaller one.
result Weight Squeezing outperforms other methods on GLUE benchmark with faster training.
Generative AI tasks analyzed for text, images, audio, video, code, and molecules.
problem What is the core question when using generative AI?
method Survey of generative model families, probabilistic framework, game-theoretic setup, post-training modifications, socially responsible considerations.
result Generative AI is a distinct machine learning task with connections to prediction, compression, and decision-making.
PAQ8 is an open source lossless data compression algorithm that currently achieves the best compression rates on many benchmarks. This report presents a detailed description of PAQ8 from a statistical machine learning perspective. It shows that it is possible to understand some of the modules of PAQ8 and use this under…
ANT learns sparse embeddings for large vocabularies efficiently.
problem Lack of scalable methods for embedding large vocabularies in neural networks.
method Anchor & Transform (ANT) algorithm that learns a small set of anchor embeddings and a sparse transformation matrix.
result ANT achieves stronger performance with fewer parameters (up to 40x compression) compared to existing methods.
This book presents a methodology and philosophy of empirical science based on large scale lossless data compression. In this view a theory is scientific if it can be used to build a data compression program, and it is valuable if it can compress a standard benchmark database to a small size, taking into account the len…
Algorithm compresses large matrices by approximating them as low rank and low precision factors.
problem Efficiently storing and processing large matrices with billions of elements.
method Randomized sketching and quantization of matrix columns to achieve low rank and low precision factorization.
result Achieves compression ratios as low as one bit per matrix coordinate while maintaining or improving performance.
We sparsify gated RNNs by simplifying their structure.
problem Improving efficiency of RNNs by reducing their complexity.
method Adjust existing sparsification techniques to gated RNNs, sparsifying preactivations of gates.
result Simplified LSTM structure improves model performance and efficiency.
This work compresses BERT into simple neural networks using unlabeled data.
problem Compression of large BERT models for practical use in downstream tasks.
method Leverage unlabeled transfer data to distill BERT into simple RNN models with hard and soft distillation.
result Simple RNN models can match or exceed BERT performance with up to 26x parameter compression.
Deep networks learn to compress high-dimensional data into lower dimensions.
problem Learning to identify and extract task-relevant variables from high-dimensional data.
method Applied state-of-the-art techniques for intrinsic dimensionality estimation.
result Neural networks first expand dimensionality in initial layers, then compress in later layers.
PCA reduces language model embeddings to improve speed and memory efficiency.
problem High-dimensional embeddings in RAG models cause scalability issues.
method Used Principal Component Analysis (PCA) to compress embeddings.
result Reducing from 3,072 to 110 dimensions provides significant speedup and memory reduction.
Funnel-Transformer reduces computation by compressing sequence data.
problem Redundant token-level representations in language processing.
method Gradually compresses sequence of hidden states to a shorter one.
result Funnel-Transformer outperforms standard Transformer with fewer FLOPs.
We consider an online version of the robust Principle Component Analysis (PCA), which arises naturally in time-varying source separations such as video foreground-background separation. This paper proposes a compressive online robust PCA with prior information for recursively separating a sequences of frames into spars…
PaGoDA reduces diffusion model training costs by 64x.
problem Diffusion models are computationally expensive during training.
method Three-stage pipeline: downsampled training, distillation, progressive super-resolution.
result PaGoDA achieves state-of-the-art performance with reduced training costs.
GENs model generates sparse graphs from text-based inputs, achieving high validity.
problem Efficiently modeling and generating sparse graphs with unique and valid structures.
method RNN-based GENs model trained with an examination mechanism to predict graph characters.
result Moderate to high validity achieved in LGI strings for sparse graph generation.
We develop embeddings for nonlinear subspaces preserving vector norms.
problem Preserving vector norms in nonlinear subspaces.
method Low-distortion embeddings for subspaces under nonlinear transformations.
result First low-distortion embeddings for a wide class of nonlinear functions.
Improves document summarization by combining word embeddings and n-grams.
problem Exact word matching fails to measure semantic similarity between sentences.
method Uses deep embedding features and tf-idf features to improve sentence similarity measure; builds an improved sentence similarity graph; employs a submodular objective function; develops a Transformer-based compression model.
result Outperforms tf-idf based approach and achieves state-of-the-art performance on DUC04 dataset.
This paper explores how NLP enhances insurance data analysis.
problem Traditional insurance data limitations and need for alternative data.
method Application of NLP techniques to transform and analyze unstructured text data.
result NLP techniques improve insurance data analysis and risk assessment.
Study shows nonextendibility of warped spacelike singularities in specific spacetimes.
problem Nonextendibility of warped spacelike singularities in specific spacetimes.
method Establishes a local obstruction through integrability conditions and radial compression.
result Imply C 0 C^0 C 0 -inextendibility for the one-horizon Birmingham-Kottler family. Flexible framework compresses models using LC algorithm.
problem Efficiently compressing neural networks for resource constraints.
method Decouples learning and compression steps with alternating L and C phases.
result Compressed models maintain performance and accuracy.
Reduces multiclass and regression compression schemes to binary ones.
problem Developing efficient learning algorithms for multiclass and regression problems.
method Reduces sample compression schemes for binary classes to multiclass and regression settings.
result Establishes new compression schemes for multiclass and regression problems.
Proposes a link between randomness and compression in deep learning.
problem Improving efficiency in deep learning training.
method Introduces a novel tomographic compression framework called Dual Tomographic Compression (DTC).
result Demonstrates high correlation between learning performance and Gibbs entropy over compression ratios.
A new method classifies images using a compressed summary.
problem Classifying images efficiently.
method Compressive learning with a novel sketch function.
result Improved image classification with less data.
Unified framework for non-compressed networks from compression bounds.
problem Generalization of large deep neural networks with huge model size.
method Unified framework converting compression based bounds to non-compressed networks.
result Unified data-dependent generalization error bound with tighter evaluation.
Paper introduces a new adaptive gradient method with gradient compression for distributed training.
problem Communication overhead in distributed machine learning systems.
method Adaptive gradient method with gradient compression, scalable system BytePS-Compress.
result Convergence rate of O ( 1 / T ) \mathcal{O}(1/\sqrt{T}) O ( 1/ T ) for non-convex problems. Galen algorithm compresses neural networks for specific hardware with reduced latency.
problem Finding optimal compression policies for neural networks on specific hardware.
method Reinforcement learning using pruning and quantization to optimize inference latency.
result Compressed ResNet18 for ARM processor reduced inference latency by 80%.
The (isothermic) compressibility of lattice knots can be examined as a model of the effects of topology and geometry on the compressibility of ring polymers. In this paper, the compressibility of minimal length lattice knots in the simple cubic, face centered cubic and body centered cubic lattices are determined. Our r…
Paper compresses neural network weight-updates for image artifacts removal.
problem Efficiently compressing neural network weight-updates for image artifacts removal.
method Fine-tuning a pre-trained artifact removal network on target data with a compression objective that encourages sparse and quantized weight-updates.
result Achieves reconstruction quality comparable to traditional codecs at comparable bitrates.
HiLLoC compresses large images losslessly using VAEs.
problem Lossless compression of large color photographs.
method Fully convolutional VAE models trained on ImageNet are applied to lossless compression.
result Achieves state-of-the-art compression for full-size ImageNet images.
Neural NCD reveals LLMs don't compress well for classification.
problem The disconnect between compression and classification in neural networks.
method Developed Neural NCD to compare LLMs to classic algorithms, finding classification accuracy not correlated with compression rate.
result Classification accuracy is not predictable by compression rate alone, challenging current understanding.
Randomly chosen support makes sparse linear regression easy.
problem Sparse linear regression with random support.
method Random support selection for efficient prediction.
result Prediction error ε ε ε with N = e x t p o l y ( k , log d , 1 / ε ) N = ext{poly}(k, \log d, 1/ε) N = e x t p o l y ( k , log d , 1/ ε ) samples and e x t p o l y ( d , N ) ext{poly}(d,N) e x t p o l y ( d , N ) run-time. BDC compresses both sample size and dimensionality of large datasets.
problem Large datasets in both sample size and dimensionality.
method Two-stage framework using Decoded MMD, Reconstruction MMD, and Encoded MMD.
result BDC achieves comparable or superior performance with lower cost and higher compression rates.