For maximum-likelihood IRL, the inner-problem Hessian at a realizable optimum equals the temperature-scaled trajectory Fisher matrix, which enables a scalable sketched hypergradient method.
∂g ∂θ θ⋆(ϕ),ϕ #−1 ∂g ∂ϕ θ⋆(ϕ),ϕ . Since ∂g ∂θ = ∂2Linner ∂θ 2 , ∂g ∂ϕ = ∂2Linner ∂θ∂ϕ , we get dθ⋆ dϕ ϕ =−
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Efficient Hypergradient Descent for Inverse Reinforcement Learning
For maximum-likelihood IRL, the inner-problem Hessian at a realizable optimum equals the temperature-scaled trajectory Fisher matrix, which enables a scalable sketched hypergradient method.