Post‑hoc Reward Calibration Reduces Length Bias in LLM Alignment

Post‑hoc reward calibration removes a length bias in RLHF reward models, improving average scores by 3.11 points across 33 models on the RewardBench dataset. Read more: getnews.me/post-hoc-reward-calibrat... #rewardcalibration #llmalignment #rlhf