Tagged articles
Sequence Modeling
24 articles · Page 1 of 1
Airbnb built a Transformer‑based sequence model that encodes years of guest behavior—including long‑term bookings and short‑term browsing—to deliver timely, personalized search results, achieving up to 3.78% overall ranking improvement and significant gains in bookings and clicks.
AirbnbSearch PersonalizationSequence Modeling
0 likes · 12 min read

The article reviews MIT’s Supervised Memory Training (SMT) and its DAgger extension (DMT), which replace traditional back‑propagation through time with a Transformer‑based teacher, enabling one‑step memory supervision for RNNs, achieving parallel‑friendly training and superior long‑sequence performance on synthetic benchmarks, TinyStories and pixel‑wise image generation.
BPTTDMTRNN
0 likes · 10 min read

PRISM demonstrates that linear‑attention models can adopt a “write‑before‑think” paradigm by reconstructing the multi‑step step‑size × residual × direction iteration of Test‑Time Training, achieving Transformer‑level quality while delivering up to 174× higher throughput through parallel scan and fused kernels.
Linear attentionParallel ScanSequence Modeling
0 likes · 19 min read

The paper systematically studies Test‑Time Training (TTT) for vision, derives six design principles, and introduces ViT³—a pure TTT architecture that uses full‑batch internal training, a learning rate of 1.0, and lightweight SwiGLU‑Depthwise convolution modules, achieving state‑of‑the‑art linear‑complexity performance across classification, detection, segmentation and generation tasks.
Computer VisionLinear ComplexitySequence Modeling
0 likes · 14 min read

HyFormer, a novel hybrid Transformer framework introduced by ByteDance’s TikTok search team, integrates sequence modeling and feature interaction into a unified alternating optimization process, enhancing representation power and scaling efficiency for ultra‑long user behavior sequences and high‑dimensional heterogeneous features, leading to significant offline and online performance gains.
AIHyFormerSequence Modeling
0 likes · 12 min read

This tutorial explains the nature of sequential data, why traditional feed‑forward networks struggle with it, and how recurrent architectures such as RNN, LSTM, and GRU capture temporal dependencies, complete with mathematical foundations, training algorithms, and full PyTorch implementations for sentiment analysis, text generation, and encoder‑decoder models.
Encoder-DecoderGRULSTM
0 likes · 57 min read

Lead–LagNet addresses three key limitations of existing graph neural networks for multivariate time‑series forecasting—loss of fine‑grained temporal detail, shared weight assumptions, and reduced interpretability—by introducing a sequence preprocessor with a global influence separator and subsequence detector, a subsequence dependency encoder, and a decoupled message‑passing mechanism, achieving superior performance on synthetic benchmarks and S&P 500 market data.
Financial Market PredictionLead‑Lag DependencyLead–LagNet
0 likes · 13 min read

Google's Titans architecture and the MIRAS theoretical framework introduce a deep neural memory that lets large language models learn in real time, retain surprising information, and handle context windows of up to two million tokens, outperforming existing Transformers and linear RNNs on a range of benchmarks.
AI MemoryMIRAS frameworkSequence Modeling
0 likes · 10 min read

The paper introduces ERDM, an enhanced rolling diffusion model that integrates progressive noise scheduling and time‑loss weighting from EDM, demonstrates superior CRPS scores on Navier‑Stokes and ERA5 mid‑term weather forecasts, and achieves comparable accuracy with far lower computational cost.
AIERDMNeurIPS 2025
0 likes · 14 min read


Qborfy AIAug 7, 2025 · Artificial Intelligence
This article explains how recurrent neural networks (RNNs) add memory to neural models, details the gate mechanisms of LSTM and GRU, compares their structures and parameter counts, and illustrates their use in speech recognition, translation, stock prediction, and video generation, while highlighting practical insights and energy considerations.
AIGRULSTM
0 likes · 5 min read

This article analyzes the fundamental differences between state space models (SSM) and Transformer architectures, highlighting their three core components, training efficiency, memory handling, tokenization impact, and empirical performance trade‑offs, and argues why SSMs can outperform Transformers on many sequence tasks.
AI ArchitectureSequence ModelingTransformers
0 likes · 19 min read

This article details the design of a community‑search pipeline that leverages long‑term user interaction sequences for CTR/CVR prediction, describes the global, online and offline architectures, enumerates the major performance and consistency challenges encountered, and presents the practical optimizations and future directions adopted to achieve reliable, high‑throughput sequence modeling.
Data ConsistencySequence Modelingai-optimization
0 likes · 12 min read


Amap TechJun 30, 2025 · Artificial Intelligence
SeqGrowGraph introduces a novel chain-of-graph expansion framework that incrementally builds lane topology graphs using a Transformer-based autoregressive model, achieving state‑of‑the‑art performance on large autonomous‑driving datasets such as nuScenes and Argoverse 2 by accurately modeling complex road structures.
Autonomous DrivingComputer VisionSequence Modeling
0 likes · 10 min read

This article presents a comprehensive solution for heterogeneous long‑behavior sequence modeling in advertising recommendation, introducing the TIN backbone, Disentangled Side‑Info TIN, Stacked TIN, and Target‑aware SASRec, along with platform‑level optimizations that enable million‑scale sequences while delivering significant online performance gains.
AdvertisingSequence ModelingTransformer
0 likes · 15 min read


NewBeeNLPMar 4, 2024 · Artificial Intelligence
This article presents a GitHub‑hosted collection of 25 recent research papers on Mamba and its variants, summarizing each work’s core contributions across sequence modeling, vision, medical imaging, graph analysis, and multimodal tasks, and highlighting their performance gains over prior methods.
Computer VisionMambaSequence Modeling
0 likes · 13 min read

The paper proposes a dual‑interest decomposition head‑attention model that uses a feedback‑aware encoding layer, a factorized head attention mechanism, and separate positive/negative interest towers to improve sequence recommendation performance on short‑video and e‑commerce datasets.
AIFeedbackSequence Modeling
0 likes · 8 min read

This article introduces recurrent neural networks (RNNs) and their ability to handle sequential data, explains the limitations of vanilla RNNs, presents the LSTM architecture with its gates, and provides complete Keras code for data loading, model building, and training both vanilla RNN and LSTM models.
KerasLSTMRNN
0 likes · 5 min read

This article reviews the evolution of behavior‑sequence modeling methods—from pooling and target‑attention to RNN, capsule, transformer, and graph neural networks—explains their industrial relevance, and demonstrates how to quickly apply these techniques in the EasyRec framework with practical configuration examples.
DINEasyRecSequence Modeling
0 likes · 21 min read

This article presents a comprehensive overview of QQ Music's recall pipeline, detailing business characteristics, challenges such as noisy user behavior and cold‑start, and four major solutions—including knowledge‑graph‑enhanced recall, sequence‑based and multi‑interest modeling, audio‑based recall, and federated learning—along with practical insights and Q&A.
Audio EmbeddingSequence Modelingknowledge graph
0 likes · 19 min read

This article details the technical evolution of Kuaishou's short‑video recommendation pipeline, focusing on sequence re‑ranking, multi‑content mixing, and on‑device re‑ranking, and explains how transformer‑based models, generator‑evaluator frameworks, and reinforcement‑learning strategies are employed to maximize overall sequence value, user engagement, and revenue.
KuaishouOn-device InferenceRe-ranking
0 likes · 15 min read


58 TechApr 12, 2021 · Artificial Intelligence
This article presents the challenges of large‑scale home‑page recommendation at 58.com, describes how behavior‑sequence models such as DIN, DIEN and Transformer are applied and evolved into double‑channel and multi‑channel deep interest architectures, and details offline and online performance optimizations that yielded significant gains in click‑through and conversion rates.
AISequence Modelinglarge-scale systems
0 likes · 19 min read

User behavior sequence modeling, crucial for search and recommendation advertising ranking, has evolved from simple pooling to attention, RNN, capsule, and Transformer architectures, with industrial applications across e‑commerce, social, video, and music platforms, and future directions include time‑aware, multi‑dimensional, and self‑supervised approaches.
AttentionSequence ModelingTransformer
0 likes · 24 min read

This article explains how LSTM networks overcome RNN limitations by using input, forget, and output gates with sigmoid and tanh activations, describes the core update equations, discusses alternative activation functions and hard‑gate variants, and provides references for deeper study.
LSTMRNNSequence Modeling
0 likes · 10 min read

The paper introduces LSTM‑Jump, a reinforcement‑learning‑trained LSTM variant that can dynamically skip irrelevant tokens, achieving up to six‑fold speed‑ups over standard sequential LSTMs while maintaining or improving accuracy on various NLP tasks such as sentiment analysis, document classification, and question answering.
LSTMNLPReinforcement Learning
0 likes · 7 min read
