Early-Signal Detector

method · first seen 15 Oct 2026

LLM-guided reward design

Also: language model reward generation

Using large language models to generate or shape reward functions and subtask plans for reinforcement learning.

Mentions per week

0119 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Earliest sources

Latest sources

Mentioned together with

potential-based reward shapingfraud detectionreject inference