LARA enables safe, constraint‑aware inference for frozen language models
The authors introduce Lagrangian Reward Augmentation (LARA), a framework that injects safety constraints into a frozen model’s decoding via a calibrated dual variable. By turning constraint handling into a one‑dimensional convex problem, LARA improves the helpfulness‑harmlessness trade‑off for both sequence‑ and token‑level inference methods, rivaling finetuning‑based alignment.