Early-Signal Detector

method · first seen 4 Oct 2026

off-policy evaluation

Also: OPE

Estimating the value of a target policy using data logged by a different policy.

Mentions per week

0719 May11 Aug10 NovPapersCodeDiscussion

PapersCodeDiscussion

Rank 99 on 16 Nov 2026 · 10 mentions in 28 days · 10 independent groups · papers · first seen 6 weeks ago · acceleration +1.71

Earliest sources

Latest sources

Mentioned together with

doubly robust estimationprior-fitted networkfitted Q-evaluationpartially observable Markov decision processbest-of-N samplingcontextual banditaugmented inverse propensity weightingPAC-Bayes boundsdiscrete action codesoffline reinforcement learning