Multi-task bilevel optimization with general lower-level convexity is recast as equality-constrained multi-objective optimization and solved by a weighted Chebyshev penalty algorithm achieving O(S T^{-1/2}) convergence to KKT-based Pareto stationarity.
Recently, (Zhang et al., 2026), for the first time in the literature, investigates the Pareto front exploration, yet their approach requires the restrictive LLSC condition
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Tale of Two Problems: Multi-Task Bilevel Learning Meets Equality Constrained Multi-Objective Optimization
Multi-task bilevel optimization with general lower-level convexity is recast as equality-constrained multi-objective optimization and solved by a weighted Chebyshev penalty algorithm achieving O(S T^{-1/2}) convergence to KKT-based Pareto stationarity.