Residual SwiGLU Deep MLP achieving 88.52% ROC-AUC, 89.50% PR-AUC, and 79.99% accuracy on the full 11M-row HIGGS dataset, after benchmarking multiple models including LightGBM, XGBoost, autoencoder-enhanced trees, and physics-engineered feature baselines.

BERT-based intent classifier achieving 98.56% accuracy, and DistilBERT-based slot filling model reaching 94.5% entity-level F1 on the SNIPS 2018 dataset. Few-shot experiments also reached 97.84% accuracy with 70 samples per intent. The full pipeline includes preprocessing, leakage-safe splits, tokenization, training, and evaluation.