Unverified50% confidenceFactExact time
Under the premises of smooth stochastic nonconvexity, a global lower bound for the population objective, and a second-moment bound on the sample Lipschitz constant, the paper gives a scalar oracle lower bound of Ω(d L_0^2 min{ΔH, L_0^2} / ε^4), which is linear in d and ε^{-4} in accuracy.
1
Sources
50%
Confidence
Long-term
Relevance
10/15/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe Stein identity states that for a function φ that decays sufficiently fast on the boundary, when the probability density p(x) is sufficiently smooth on R^d, E_p[∇_x log p(x)·φ(x) + ∇_x·φ(x)] = 0.68% similarUnverifiedThe query complexity given in the paper for finding a Goldstein stationary point under first-order access (stochastic gradient queries) is O(δ⁻¹ε⁻³), where δ is the Goldstein radius and ε is the stationarity tolerance.65% similarUnverifiedIn the non-convex, high-dimensional, small-accuracy regime, the reasonable goal for zeroth-order algorithms is linear in d and close to ε^{-4} in accuracy, rather than pursuing sublinear dimension.64% similarUnverifiedThe main result of the paper is that Q-Learning with zero initialization and constant step size can, with high probability, reduce the sup-norm error to within ε using Õ(H^4/(μ_min·ε^2)+h_star) state transitions (for 0<ε≤1).64% similarUnverifiedBased on Markov random product bounds, the paper provides conditional moment contraction results for RETD under a constant step size.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1037304API
curl https://kongchang.com/api/v1/knowledge/claims/1037304MCP
get_claim(id=1037304)