1: \begin{abstract}

2: Estimation of the complete distribution of a random variable is

3: a useful primitive for both manual and automated decision making.

4: This problem has received extensive attention in the i.i.d. setting,

5: but the arbitrary data dependent setting remains largely unaddressed.

6: Consistent with known impossibility results, we present computationally

7: felicitous time-uniform and value-uniform bounds on the CDF of the running

8: averaged conditional distribution of a real-valued random variable which

9: are always valid and sometimes trivial, along with an instance-dependent

10: convergence guarantee.  The importance-weighted extension is appropriate for

11: estimating complete counterfactual distributions of rewards given controlled

12: experimentation data exhaust, e.g., from an A/B test or a contextual bandit.

13: \end{abstract}

14: