Tag
This paper proposes separating generation from selection in explainable-recommendation systems to reduce serving costs, using a frozen candidate pool of explanations and a small CPU-resident selector. It benchmarks offline-pool selectors and finds that pairwise learning-to-rank outperforms single-action RL formulations like PPO, GRPO, and DPO in terms of F1 scores.
A blog post from TensorZero argues that even very noisy LLM evaluators can be useful for offline agent selection and improvement, as noise averages out over many samples to reliably rank agents.