[NeurIPS 2025] Source code of "STaRFormer: Semi-Supervised Task-Informed Representation Learning via Dynamic Attention-Based Regional Masking for Sequential Data"
A viewer whose perception evolves with each image—stateful, memory-carrying VLM reflections across a gallery.
Malware classification framework leveraging dynamic API-call sequences. Explores multiple Deep Learning architectures including Frequency-based FFNNs, Recurrent Neural Networks (GRU, BiLSTM) for sequential analysis, and Graph Neural Networks (GraphSAGE, GCN) for structural behavioral modeling. Focuses on feature extraction and sequence embedding.
Sequential forecasting pipeline leveraging Gated Recurrent Units (GRU) to model long-term dependencies in ad performance data. A deep learning baseline comparing RNN architectures against TCNs for temporal dynamics.
Project Page of [NeurIPS-2025] "STaRFormer: Semi-Supervised Task-Informed Representation Learning via Dynamic Attention-Based Regional Masking for Sequential Data"
A research-oriented implementation of sequential generative models for temporal data and NLU.
> 🛡️ Lightweight LSTM-based system log anomaly detection using sequential event modeling, with an interactive Streamlit dashboard for real-time analysis and visualization.
To associate your repository with the sequential-modeling topic, visit your repo's landing page and select "manage topics."