REVIEW 4 cited by
On Penalty Methods for Nonconvex Bilevel Optimization and First-Order Stochastic Approximation
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
abstract
In this work, we study first-order algorithms for solving Bilevel Optimization (BO) where the objective functions are smooth but possibly nonconvex in both levels and the variables are restricted to closed convex sets. As a first step, we study the landscape of BO through the lens of penalty methods, in which the upper- and lower-level objectives are combined in a weighted sum with penalty parameter $\sigma > 0$. In particular, we establish a strong connection between the penalty function and the hyper-objective by explicitly characterizing the conditions under which the values and derivatives of the two must be $O(\sigma)$-close. A by-product of our analysis is the explicit formula for the gradient of hyper-objective when the lower-level problem has multiple solutions under minimal conditions, which could be of independent interest. Next, viewing the penalty formulation as $O(\sigma)$-approximation of the original BO, we propose first-order algorithms that find an $\epsilon$-stationary solution by optimizing the penalty formulation with $\sigma = O(\epsilon)$. When the perturbed lower-level problem uniformly satisfies the small-error proximal error-bound (EB) condition, we propose a first-order algorithm that converges to an $\epsilon$-stationary point of the penalty function, using in total $O(\epsilon^{-3})$ and $O(\epsilon^{-7})$ accesses to first-order (stochastic) gradient oracles when the oracle is deterministic and oracles are noisy, respectively. Under an additional assumption on stochastic oracles, we show that the algorithm can be implemented in a fully {\it single-loop} manner, i.e., with $O(1)$ samples per iteration, and achieves the improved oracle-complexity of $O(\epsilon^{-3})$ and $O(\epsilon^{-5})$, respectively.
Forward citations
Cited by 4 Pith papers
-
Stochastic Dynamic Barrier Perturbed Gradient Methods for Nonconvex Simple Bilevel Optimization
SDBPG and its penalty/variance-reduced variants achieve the first explicit (ε_f, ε_g)-stationarity rates for stochastic nonconvex simple bilevel problems by regularizing unbounded dual multipliers.
-
Sharper Analysis of Single-Loop Methods for Bilevel Optimization
Decoupled-norm analysis improves single-loop AID to O(κ⁵/K) and shows single-loop ITD's asymptotic error is exactly O(κ²), matching the known lower bound.
-
Nonconvex Decentralized Stochastic Bilevel Optimization under Heavy-Tailed Noise
The paper introduces D-NSVRGDA, a decentralized normalized variance-reduced method for nonconvex bilevel optimization, and proves the first convergence rate under heavy-tailed noise without gradient clipping.
-
Beyond Two-Stage Training: Cooperative SFT and RL for LLM Reasoning
BRIDGE couples SFT and RL via bilevel optimization plus a cooperative-gain LoRA objective and reports consistent math-reasoning gains over cold-start and mixing baselines.
Discussion (0). Continue with ORCID to comment.