Fix agentic Pareto frontier intersections - #674
Merged
Conversation
Compute each selected-axis Pareto frontier over all eligible points before intersecting it with the E2E Normalized Interactivity winner set. Add MiniMax M3 regression cases for the E2E latency and TTFT zig-zags, keep GPU comparison rooflines x-sorted, and update the bilingual Optimal Only explanation. 中文：修复（智能体）：对规范化与当前横轴的 Pareto 前沿取交集。先基于全部有效点计算当前横轴的 Pareto 前沿，再与端到端归一化交互性优胜点集合取交集；新增 MiniMax M3 端到端延迟与 TTFT 折线异常的回归用例，确保 GPU 对比曲线按 x 排序，并同步更新中英文“仅最优”说明。
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
cquil11
changed the title
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Root cause
PR #638 re-plotted the exact normalized-interactivity winner set on E2E latency, Interactivity, and TTFT without another local Pareto test. A point that won on the normalized axis could therefore remain connected after the x-axis changed even when another point dominated it on E2E latency or TTFT, producing the zig-zag curves visible in the MiniMax M3 run.
The corrected order is: compute Pareto over all eligible points in the displayed coordinate system, then retain only members of the normalized North Star frontier. This is the mathematical intersection of both frontier sets and prevents non-canonical dominators from being ignored.
Validation