IQ-Learn#

Reference PDF: pending known-truth migration.

IQ-Learn parameterizes soft Q-values directly and recovers a Bellman-implied reward as a byproduct. It is part of the 12-estimator known-truth validation suite.

This RTD page is intentionally short while the dedicated IQ-Learn TeX/PDF tutorial is migrated to the shared synthetic DGP. No real-data estimation examples are published on RTD.

Migration Target#

  • PDF source: papers/econirl_package/primers/iq_learn/iq_learn.tex

  • Result generator: papers/econirl_package/primers/iq_learn/iq_learn_run.py

  • Shared DGP harness: experiments/known_truth.py