KV‑Cache Rewards Cut Multi‑Agent LLM Scoring Cost 5,000×
KV-PRM reads the LLM's KV cache instead of re‑encoding whole trajectories, turning reward scoring from quadratic to linear time. The trick slashes FLOPs by up to 5,000×, cuts latency 37× and memory 34×, while matching or beating text‑based PRMs on MATH, GSM8K and AIME.