Natural Language Processing at Yale
References Improve LLM Alignment in Non-Verifiable Domains
ANCHOR: Branch-Point Data Generation for GUI Agents