Signals 4 · free daily AI digest

Tail-Influence Sampling for CVaR Policy Evaluation

Paper recorded by Signals 4 on 2026-09-29 in cs.LG. Abstract reproduced from arXiv; link to the original below.

Published 2026-09-29 on arXiv · recorded by Signals 4 on 2026-09-30

Category: cs.LG · 机器学习 · first seen 2026-09-30

Abstract

Policies with similar mean returns can differ sharply in rare failures, yet estimating lower-tail conditional value-at-risk (CVaR) accurately can require many costly rollouts. When different conditional components of a stochastic workflow can be queried separately, we ask how to allocate a fixed evaluation budget to estimate a fixed policy's CVaR most accurately. We derive a tail influence for eac

Read on arXiv →

#9 most recent of 334 cs.LG papers we have recorded · ↑ newer: Explore Broadly, Reason Sharply: Push Small Models toward the Frontier · ↓ older: Probe-Space Preconditioning for Fast and Stable Zero-Order Training
Cite this page: Tail-Influence Sampling for CVaR Policy Evaluation: the #9 most recent of 334 cs.LG papers we have recorded (as of 2026-09-29). Source: Signals 4 (Signals API) — https://data.jiangzhang.ca/signals4/t/papers/tail-influence-sampling-for-cvar-policy-evaluation.html
Free to quote with attribution to “Signals 4 (Signals API)”. Machine-readable: papers.json
Related: More cs.LG papers · arXiv signals · All papers · Today in AI
Get 4 AI signals a day by email — free.
Subscribe free → See all plans →
Get 4 AI signals a day by email — free
All models · All repos · By company · Daily editions