- Rating
- 1
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
The paper proposes clipping the lower-level stochastic gradient in two-timescale bilevel optimization under heavy-tailed noise. The clipping threshold is set from a rolling quantile of recent gradient norms. The analysis studies a variant with a deterministic, polynomially growing threshold under noise with bounded -th moment, . Experiments cover synthetic problems, two vision tasks, a stochastic zero-sum game, and offline actor-critic.
Critical strengths
- The paper addresses a relevant problem. Large gradient outliers can make bilevel optimization unstable, and using recent gradient information to adjust clipping is a reasonable idea.
- The experiments cover several settings and include a fixed-threshold variant. This helps assess whether adapting the threshold matters in practice.
Critical weaknesses
-
Assumption 4.2 bounds the noise around the true gradient. However, Theorem 4.4 uses this noise bound to control the expected squared norm of the clipped gradient itself. These are different quantities. For example, take a deterministic gradient equal to 2, with and . The noise assumption is satisfied because there is no noise. If the clipping threshold is sufficiently large, the gradient remains unchanged, so its squared norm is 4. The theorem nevertheless bounds it by 1. The appendix introduces a term involving the true gradient but then drops it. This term needs to be handled explicitly, because the lower-level tracking proof relies on the resulting bound.
-
The method requires the outer step to become smaller relative to the lower-level step , so that the lower-level iterate can track the changing solution. However, the theorem chooses This gives , which grows rather than tending to zero. With this schedule, the accumulated tracking-error contribution in the final bound also grows instead of vanishing. The proof therefore does not establish the claimed convergence rate. Please clarify the intended schedule and redo the rate calculation, since swapping the exponents also changes the balance of the other error terms.
Questions for authors
-
Can you give a valid clipping bound that accounts for the true gradient and explain how it changes the analysis?
-
Which step-size schedule is intended, and does the corrected analysis still give the claimed rate?