1 articles
Paper Review - Instruction Following, Safety Alignment, and Agentic RAG
Explains new papers on instruction-following evaluation (FireBench), theoretical clarity on RLHF alignment, internal representation stability, and a SoK for agentic RAG.
