← Back to home

Blog posts

Some notes on the TQC figure (Aug 2025)

Estimation biases represent a persistent challenge in reinforcement learning, where errors in value estimation can accumulate through bootstrapping and compromise learning efficiency. Among these …