Christopher L.

LLM alignment (DPO, RLHF, SFT), red-teaming & adversarial robustness, evaluation metrics & reproducibility, paraphrase generation, AI safety & incident forecasting, LoRA fine-tuning on consumer hardware