Last-Iterate Rate for Anchored Gradient Descent-Ascent

For smooth convex-concave min-max problems, can anchored gradient descent-ascent be scheduled so that its exact last-iterate squared-gradient residual is $O(1/t)$, closing the gap left by the 2019 analysis?

Exact FrontierDelta

Prior state unknown→proved

Scope and record

Occurred: Apr 4, 2026

Delta type: SOURCE CLAIM

Assumptions: VibeMathed verification: lean-verified. Publication: preprint. AI contribution: ai-discovered. Imported under CC BY 4.0.

Canonical aliases: Last-Iterate Rate for Anchored Gradient Descent-Ascent · Anchored GDA rate

Confidence: Not scored

Registry verification: lean verified · preprint · resolved

Open the source record ↗

Attribution

VibeMathed
registry · event recorded by

AlphaProof Nexus
model · ai model contributor · Google DeepMind

Lineage and corrections

This event attributed to AlphaProof Nexus

Act on this frontier

Verify, challenge, or extend the result.

Help verify Open canonical problem