Long Short-Term Memory Networks

Also known as LSTM

LSTMs process sequences step by step, using input, forget, and output gates to control what information enters, persists in, and leaves a memory state. Researchers apply them to time-series forecasting, text and speech tasks, sensor-based prediction, and control, often comparing them with simpler statistical models, convolutional networks, Transformers, and state-space models. Common extensions include bidirectional and attention-based LSTMs, ensembles, hybrid architectures, and methods that add physical constraints or estimate uncertainty. LSTMs remain useful for many sequence tasks, but their accuracy depends on the data and evaluation setting; limited parallelism and difficulty learning long-range dependencies compete with their simplicity and effectiveness on moderate-length sequences.

Latest papers

41 papers7 papers in the last four weeks, against 2 the four weeks before. 0.0% of all new papers.

Memory or a Clock? Auditing a Memory Advantage in XLand-MiniGrid with Seeds, Maps and Evaluation Noise Apart
Oct 24, 2026Rodrigo Reveron-FonsecaReinforcement Learning Memory-Augmented RL
Tecnológico de Monterrey
StarRTT: An End-to-End Real-Time Framework for Blind Zone-Free RTT Prediction in Starlink LEO Satellite Networks
Oct 23, 2026Jaeyong Lee, Seongjin Jeong, Yusang Park +3LEO Satellite Networks Spatiotemporal Forecasting
School of Electrical Engineering, Korea University, Seoul 02841, South Korea
Gate Initialisation versus Recurrent Dropout and Zoneout in Learning Long Dependencies: A Controlled Study of LSTM and GRU Networks on Synthetic Tasks
Oct 21, 2026Oleksandr Kholodniak, Oleksandr ProkhorovRecurrent Neural Networks Long Short-Term Memory Networks
National Aerospace University “Kharkiv Aviation Institute”, Kharkiv, Ukraine
Learning volatility-adaptive forgetting via GARCH-modulated LSTM for stock price forecasting
Oct 17, 2026Xin-Rui Lu, Da-Qing ZhangStock Price Forecasting Time Series Forecasting
School of Science, University of Science and Technology Liaoning, Anshan, 114051, Liaoning, PR China
Sequence-to-Sequence Neural Surrogate Modeling of Transient Responses in High-Speed Links with Variable Design Parameters
Oct 16, 2026Julian Withöft, Shayan Ghanbari, Werner John +3Neural Surrogate Modeling Sequence-to-Sequence Models
Information Processing Lab, Faculty for Electrical Engineering and Information Technology, TU Dortmund, Germany · Pyramide2525, Paderborn, Germany and the Information Processing Lab, Faculty for Electrical Engineering and Information Technology, TU Dortmund, Germany · EMC Technology Center Paderborn, Zuken GmbH, Paderborn, Germany
The RNN Homunculus: Learning to Gate Long-Term Memory
Oct 9, 2026Zeyu Fan, Alexandru-Dragos Manolache, Jan van GemertLong-Range Dependency Modeling Recurrent Neural Networks
Delft University of Technology, Mekelweg 5, 2628 CD Delft, The Netherlands
Signal Simplification Is Not Predictive Simplification: Diagnosing Residual Neural Forecasting in Short-Horizon Volatility
Oct 2, 2026Bingqi Lian, Linfeng Cheng, Mei Lu +1Time Series Forecasting Residual Learning
Science Academy, College of Computer, Mathematical, and Natural Sciences, University of Maryland, College Park, MD 20742, USA · Department of Electrical and Computer Engineering, University of Maryland, College Park, MD 20742, USA
VertiFuseX: Generalizable Financial Forecasting via Multi-Stream Temporal Fusion
Sep 11, 2026Aashish Bohra, Vivek VijayFinancial Forecasting Non-Stationary Time Series Forecasting
Department of Computer Science and Engineering, Indian Institute of Technology Jodhpur, Jodhpur, Rajasthan, India · Department of Mathematics, Indian Institute of Technology Jodhpur, Jodhour, Inida
X-RACE: XAI-assisted Recurrent neural network Attribution for Channel Estimation
Sep 10, 2026Abdul Karim Gizzini, Yahia MedjahdiEfficient Neural Network Inference Explainable Artificial Intelligence
University of Paris-Est Cr´eteil (UPEC), LISSI/TincNET, F-94400, Vitry-sur-Seine, France. · IMT Nord Europe, Institut Mines T´el´ecom, Centre for Digital Systems, F-59653 Villeneuve d’Ascq, France.
Multi-Step Forecasting of Grape Berry Temperature based on LSTM Model with Feed-Forward Attention
Aug 29, 2026Srikanth Gorthi, L. G. Divyanth, Dattatray Bhalekar +2Time Series Forecasting Long Short-Term Memory Networks
Center for Precision and Automated Agricultural Systems & Department of Biological Systems Engineering, Washington State University, Prosser, WA, USA · AgWeatherNet, Washington State University, Prosser, WA, USA · Department of Biological & Environmental Engineering, Cornell University, Ithaca, NY, USA +1
Directional Contextual Representations for Dependency Relations: Why Cross-Direction Pairing Fails
Aug 21, 2026Sai Krishna Arthanari, JaeHyeong Chang, Chengzhe Sun +1Representation Learning Long Short-Term Memory Networks
Institute for Artificial Intelligence and Data Science (IAD), University at Buffalo Buffalo, NY, USA
A Multi-Timestep LSTM Ensemble regressor for Enhanced Short-Term Runoff Prediction
Aug 14, 2026Hamid Saadatfar, AmirHossein Eshghi, Behnaz BehdaniTime Series Forecasting Long Short-Term Memory Networks
Department of Computer Engineering, Faculty of Electrical and Computer Engineering, University of Birjand, Birjand, Iran · Department of Computer Science, Clemson University, Clemson, SC, USA
XGBoost "is all you need": the case of forecasting transmitted heat energy in District Heating Systems
Aug 11, 2026Milan ZdravkovićGradient-Boosted Decision Trees Energy-Efficient ML
Faculty of Mechanical Engineering University of Niš ul. Aleksandra Medvedeva 14, 18000 Niš, Serbia
MambaLSTM: A Spatio-Temporal Framework for Enhanced Traffic Accident Risk Prediction
Jul 20, 2026Zhen Yu, Yachao Yuan, Zixiang Peng +2Traffic Accident Anticipation Long Short-Term Memory Networks
School of Future Science and Engineering, Soochow University, Suzhou, Jiangsu, China · Key Laboratory of General Artificial Intelligence and Large Models in Provincial Universities, Soochow University, Suzhou, Jiangsu, China · College of Computing and Intelligent Systems, University of Khorfakkan, Sharjah, UAE
The Orthogonalized Read Is a Removable Training Scaffold for Recurrent Memory
Jul 2, 2026Keston Aquino-MichaelsLong Short-Term Memory Networks Associative Memory
No Way Labs
LSTM Variants for Chaotic Dynamical Systems: An Empirical Study on the Lorenz Attractor
Jun 21, 2026Ruslan GokhmanTemporal Convolutional Networks Chaotic Dynamical Systems
Yeshiva University
David vs. Goliath in Next Activity Prediction: Argmax vs. LSTM, Transformer, and LLM
Jun 14, 2026Hans Weytjens, Ingo WeberLLM Evaluation Predictive Process Monitoring
Technical University of Munich, School of CIT, Germany · Fraunhofer Gesellschaft, Munich, Germany
Towards Critical Branching Mechanism in Recurrent Neural Networks
Jun 9, 2026Feixiang Ren, Ling FengRecurrent Neural Networks Long Short-Term Memory Networks
Department of Physics, National University of Singapore Singapore · Institute of High Performance Computing Agency for Science, Technology and Research (A*STAR) Singapore
Physics-Informed Machine Learning for Short-Term Flood Prediction
Jun 2, 2026Tewodros Syum Gebre, Jagrati Talreja, Leila Hashemi-BeniLong Short-Term Memory Networks Physics-Informed ML
North Carolina A&T State University · Greensboro, NC 27407, USA
Few-Shot Prediction for Pulsar Noise with Long Short-Term Memory Network
Jun 2, 2026Qingye Tang, Dechao An, Haoran Peng +1Few-Shot Learning Long Short-Term Memory Networks
Sichuan University, College of Computer Science · Sichuan University, College of Physics · Sichuan University
Speech Emotion Recognition using Attention-based LSTM-Network with Residual Connection
Jun 2, 2026Daniil Krasnoproshin, Maxim VashkevichSpeech Emotion Recognition Long Short-Term Memory Networks
Embedded System Department of Belarusian State University of Informatics and Radioelectronics 6, P. Brovky st., 220013, Minsk, Belarus
Evaluating Transformer and LSTM Frameworks for Prediction in Ungauged Basins
Jun 1, 2026Taye Akinrele, James Halgren, Noorbakhsh Amiri Golilarz +2Streamflow Forecasting Transformer
Dept. of Computer Science, The University of Alabama, Tuscaloosa, USA · Alabama Water Institute, USA
Comparative Analysis of Liquid Neural Networks and LSTM for Sequential Pattern Recognition: Robustness, Efficiency, and Clinical Utility
May 26, 2026Ye Kyaw Thu, Thazin Myint Oo, Thepchai SupnithiEfficient Neural Network Inference Liquid Neural Networks
National Electronics and Computer Technology Center (NECTEC), Pathumthani, Thailand · Language Understanding Lab., Yangon, Myanmar
An Objective Performance Evaluation of the LSTM Networks in Time Series Classification
May 19, 2026Sooraj Sunil, Balakumar BalasingamTime Series Classification Recurrent Neural Networks
Electrical and Computer Engineering University of Windsor Windsor, ON, Canada
Parallel Recursive LSTM
May 16, 2026Tristan Gaudreault, Yongyi MaoRecurrent Neural Networks Language Modeling
School of Electrical Engineering and Computer Science University of Ottawa Ottawa, ON
Ti-iLSTM: A TinyDL Approach for Logic-Level Anomaly Detection in Industrial Water Treatment Systems
May 15, 2026Mandar Joshi, Farzana Zahid, Judy Bowen +3Efficient Neural Network Inference Industrial Anomaly Detection
University of Waikato, Hamilton, New Zealand · Auckland University of Technology, Auckland, New Zealand · Aalto University, Espoo, Finland
A Hybrid Tucker-LSTM Tensor Network Model for SOC Prediction in Electric Vehicles
May 13, 2026Han Wang, Ying Wang, Bing WangLong Short-Term Memory Networks
College of Computer and Information Science Southwest University Chongqing, China · School of Culture Tourism Chongqing Vocational College of Culture and Arts Chongqing, China · Digital Intelligence Center China Automotive Engineering Research Institute Co., Ltd. Chongqing, China
Dual-Temporal LSTM with Hybrid Attention for Airline Passenger Load Factor Forecasting: Integrating Intra-Flight and Inter-Flight Booking Dynamics
May 12, 2026ASM Nazrul Islam, Md. Hasanul Kabir, Md. Liakot Ali +1Time Series Forecasting Long Short-Term Memory Networks
Institute of Information and Communication Technology, Bangladesh University of Engineering and Technology, Dhaka, Bangladesh · Islamic University of Technology, Dhaka, Bangladesh
Neural-Actuarial Longevity Forecasting: Anchoring LSTMs for Explainable Risk Management
May 7, 2026Davide RindoriLong-Term Time Series Forecasting AI Risk Management
SAV Actuarial Candidate, ETH Zürich · PhD in Physics, University of Florence (2021)
Sentiment Analysis of Indonesian Spotify Reviews Using Machine Learning and BiLSTM
May 5, 2026Uliano Wilyam Purba, Andre Hadiman Rotua Parhusip, Sahid Maulana +3Sentiment Analysis Multiclass Classification
Department of Data Science Faculty of Science Sumatera Institute of Technology Lampung, Indonesia · Department of Informatics Engineering Faculty of Industrial Technology Sumatra Institute of Technology

Oct 24, 2026J/K move · Enter open · S save

Memory or a Clock? Auditing a Memory Advantage in XLand-MiniGrid with Seeds, Maps and Evaluation Noise Apart

Rodrigo Reveron-Fonseca

Tecnológico de Monterrey

Reinforcement Learning Memory-Augmented RL Partially Observable RL Long Short-Term Memory Networks

Open paper AskPDF arXiv

On a partially observable XLand-MiniGrid task, an LSTM agent beats a memory-free MLP by 25 percentage points. The gap looks like a clear case of memory. We ask two questions about it. First, how much of a measured gap is noise? Separating training seed, map and evaluation randomness in 60 runs, each re-evaluated three times, we find that about ten paired seeds detect a 10-point gap in final success. The binary "task acquired" verdict, a common summary of learning curves, is 76% evaluation noise and would need about a hundred seeds for the same gap. Second, is the gap memory? On a sealed panel of new maps, opened once, a memory-free MLP of the LSTM's size comes within 3.3 points of the LSTM when it is given the time step, and gains no measurable benefit from its width without it. With the time step, an MLP a quarter of the LSTM's size, and the LSTM with its state erased at every step, also come within about 5 points. A linear probe reads the time step from the trained LSTM's state. On this task, most of the apparent memory advantage is an advantage of knowing the time, which matters because episodes are cut at a fixed limit and the reward shrinks with time. We recommend giving memory-free baselines the time step, and we release the protocol, the event-driven tooling and every checkpoint.