Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,181 papers · 148 categories

Trend · papers per month

4108191,2291,638 · Jun 202019922001200920182026
48 results for Long Sequence Learning

LTM tackles long sequence language modeling by avoiding vanishing and exploding gradients.

problem Language models struggle with long sequences due to gradient issues.
method Introduces Long Term Memory network (LTM) that scales memory and weights input, avoiding overfitting.
result LTM achieves state-of-the-art perplexity results with fewer cells than previous models.

This paper improves RNN memory capacity for long sequences through learning associative memory update rules.

problem Challenges in RNNs remembering long sequences.
method Jointly learns memory update rule with task objective and uses multiple associative memories.
result Improves memory capacity for long sequence encoding.

Reinforcement learning after next-token prediction aids in learning from diverse sequence lengths.

problem Learning from sequences of varying lengths and complexity.
method Introducing a framework to study reinforcement learning with autoregressive transformers, focusing on next-token prediction and mixture distributions of short and long sequences.
result Reinforcement learning after next-token prediction enables autoregressive transformers to generalize from long sequences, even when they are rare.

A new method backtracks through a few key past states to speed up credit assignment in long sequences.

problem Computational inefficiency of back-propagation through time for long sequences.
method Sparse attentive backtracking using learned attention mechanisms to skip connections.
result Matches or outperforms regular BPTT and truncated BPTT in tasks with long-term dependencies.

New Performer model tackles long-sequence protein modeling.

problem Challenges of training complex Transformer models for long sequences.
method Linearly scalable long-context Transformer architecture, Performer.
result Performer provides strong theoretical guarantees and is effective for protein sequence modeling.

Cloned HMMs efficiently learn variable order sequences without local minima issues.

problem Learning long-term temporal structure in sequences.
method Constrained HMMs with a sparsity structure that maps hidden states deterministically to emissions.
result Cloned HMMs can model temporal dependencies at arbitrarily long distances and recognize contexts with 'holes'.

Hrrformer uses HRR to speed up self-attention for long sequences.

problem Infeasibility of using transformers with very long sequence lengths.
method Re-cast self-attention using Holographic Reduced Representations (HRR).
result Achieves near state-of-the-art accuracy with O(THlogH)\mathcal{O}(T H \log H) time complexity and O(TH)\mathcal{O}(T H) space complexity.

Several machine learning problems arising in natural language processing can be modeled as a sequence labeling problem. We provide Gaussian process models based on pseudo-likelihood approximation to perform sequence labeling. Gaussian processes (GPs) provide a Bayesian approach to learning in a kernel based framework. …

2014-12-25abs ↗pdf ↗

TRM improves long-horizon reinforcement learning for LLMs by masking divergent sequences.

problem Long-horizon reinforcement learning for LLMs suffers from off-policy mismatch and approximation errors.
method Derives and applies trust region bounds to control divergence, proposing Trust Region Masking.
result First non-vacuous monotonic improvement guarantees for long-horizon LLM-RL.

Sparse Transformers reduce memory and time requirements for long sequence modeling.

problem Quadratic growth in memory and time with transformer sequence length.
method Sparse factorizations of attention matrix, deeper network training, attention matrix recomputation, fast attention kernels.
result Sparse Transformers can model sequences up to tens of thousands of timesteps with hundreds of layers.

New method improves deep learning model robustness and accuracy for long sequences.

problem Challenges in learning long-range sequence tasks using state-space models.
method Proposes a perturb-then-diagonalize (PTD) methodology to address ill-posed diagonalization problems in SSMs.
result Demonstrates improved robustness and accuracy of S5-PTD model on Long-Range Arena benchmark.

TRM improves long-horizon LLM RL by masking divergent sequences.

problem Long-horizon reinforcement learning with LLMs suffers from off-policy mismatch and approximation errors.
method Derives and applies trust region bounds to control divergence, proposing Trust Region Masking.
result First non-vacuous monotonic improvement guarantees for long-horizon LLM-RL.

Spacetimeformer learns spatiotemporal relationships from data alone.

problem Forecasting multivariate time series with distinct spatial relationships.
method Transformers with dynamic graph connections learning interactions between space, time, and value.
result Competitive results on various time series prediction benchmarks.

TFiLM expands convolutional models' receptive field with minimal overhead.

problem Capturing long-range dependencies in sequential data.
method A novel architectural component using a recurrent neural network to modulate convolutional model activations.
result TFiLM significantly improves learning speed and accuracy on various tasks.

S3^3Attention improves long sequence attention with smoothed skeleton sketching.

problem Quadratic complexity of vanilla Attention makes it unsuitable for long sequence tasks.
method S3^3Attention uses smoothing and matrix sketching to balance information preservation and computation.
result S3^3Attention significantly outperforms vanilla Attention and other Attention variants.

A new model improves recurrent neural networks' ability to memorize long sequences.

problem Improving recurrent neural networks' ability to memorize long sequences and extract task-relevant features.
method Proposes a Linear Memory Network with an encoding-based memorization component and a specialized training algorithm.
result Improves the final performance of recurrent neural networks when memorizing long sequences is necessary.

Neural M3 model adapts to diverse user behaviors over short and long timeframes.

problem Adapting to diverse user behaviors over short and long timeframes.
method Neural Multi-temporal-range Mixture Model (M3) combining short-term and long-term models with a learned gating mechanism.
result M3 consistently outperforms state-of-the-art sequential recommendation methods.

A new method predicts protein functions using variable-length sequences.

problem Computational methods for protein function prediction are slow and inaccurate for long sequences.
method Two feature sets: single fixed-sized segments and multi-sized segments, using bi-directional LSTM. Combined with MLDA features.
result Significant improvement in accuracy for long protein sequences.

New method maps protein sequences to embeddings encoding structural information.

problem Inferring structural properties from amino acid sequences when structures are unknown.
method Representation learning using bidirectional LSTM models with structural similarity and residue contact maps.
result Trained embeddings improve structural similarity prediction and transfer to other tasks.

Graph WaveNet models spatial-temporal graphs by learning hidden dependencies and long sequences.

problem Capturing hidden spatial dependencies and long-range temporal sequences in graphs.
method Graph WaveNet integrates adaptive dependency matrix learning and stacked dilated 1D convolution.
result Graph WaveNet outperforms existing methods on public traffic network datasets.

Neural Physicist learns physical dynamics from images.

problem Learning meaningful physical state representations and accurate state transitions from image sequences.
method Neural Physicist uses VAE for state extraction, NP for parameters, and SSM for dynamics.
result Achieves long-term predictions and identifies system degrees of freedom.

HGConv uses HRR to efficiently detect malware, outperforming existing methods.

problem Efficiently detecting malware with long sequences.
method Holographic Global Convolutional Networks (HGConv) utilizing Holographic Reduced Representations (HRR).
result Achieved state-of-the-art results on malware benchmarks.

This paper studies mutual information in sequence models and finds Transformers excel in capturing long-range dependencies.

problem Understanding the expressive power of sequence models in capturing temporal dependencies.
method Theoretical and empirical analysis of linear and nonlinear RNNs, including Transformers.
result Transformers can capture long-range mutual information more efficiently than RNNs.

SSMs have a built-in bias towards low-frequency components, which can be adjusted.

problem Frequency bias in SSMs affects their performance on long-range sequences.
method Proposed two mechanisms to tune frequency bias: scaling initialization or applying a Sobolev-norm-based filter.
result Tuning frequency bias improves SSMs' performance on long-range sequence learning tasks.

Skeinformer accelerates self-attention for long sequences with linear complexity.

problem Efficiency of Transformer models in processing long sequences.
method Matrix sketching and column sampling to reduce quadratic complexity to linear.
result Skeinformer outperforms alternatives with smaller time/space footprint.

Introduces Spectral Attention for better long-range time series forecasting.

problem Challenges in capturing long-range dependencies in time series forecasting.
method Spectral Attention mechanism that preserves temporal correlations and long-range dependencies.
result Achieves state-of-the-art results on 11 real-world time series datasets.

MusicVAE uses a hierarchical decoder to model long-term structure in music sequences.

problem Difficulty of existing recurrent VAE models in modeling long-term structure in sequential data.
method Proposes a hierarchical decoder that outputs embeddings for subsequences and uses these embeddings to generate each subsequence independently.
result Demonstrates better sampling, interpolation, and reconstruction performance than a flat baseline model.

IGLOO slices features space to process long sequences efficiently.

problem Efficiency in processing long sequences with neural networks.
method Slicing features space into non-local patches and using their relationships to build sequence representations.
result IGLOO can handle dependencies up to 20,000 steps efficiently.

D-LinOSS models learn to dissipate energy, improving performance on long-range tasks.

problem Representational limitations of LinOSS models in long-range reasoning.
method Introducing Damped Linear Oscillatory State-Space models (D-LinOSS) that learn to dissipate latent state energy on arbitrary time scales.
result D-LinOSS consistently outperforms previous LinOSS methods on long-range learning tasks, achieving faster convergence and reducing hyperparameter search space.