AI & ML
SOTA or Luck? The Winner's Curse in LLM Leaderboards
arXiv stat.ML0 views
arXiv:2611.22227v1 Announce Type: cross Abstract: Public LLM leaderboards guide model evaluation and adoption, yet the reported performance of top-ranked models can be systematically overstated. This leaderboard winner's curse arises because ranking noisy benchmark scores rewards sampling luck alongside genuine ability. The same selection effect can distort scores throughout the leaderboard, a phenomenon we call the rank-induced bias. We develop an $m$-out-of-$N$ bootstrap to correct for this bias. It assumes a fixed model pool with independent, identically distributed bounded tasks while allowing arbitrary within-task correlation among models. Under exact population tiers separated by positive gaps, we prove that the bootstrap reduces $\Theta(N^{-1/2})$ ranking bias to expected residual bias $o(N^{-1/2})$. Across 14 benchmarks from 5 benchmark families, the rank-1 adjustments range from approximately 0.09 to 3.86 percentage points. In our SWE-bench Verified snapshot, the tied leaders receive an adjustment of 0.90 percentage points. Such overstatement is large enough to make the apparent winner statistically fragile. Thus, we use the same bootstrap to further quantify the intervals of plausible rank positions for each model and group models into tiers. We establish sufficient conditions for consistent recovery of population tiers. On SWE-bench Verified, this procedure places 13 of 173 models in the leading tier. We combine the score adjustment and rank intervals in the Rank Uncertainty and Leaderboard Evaluation Report (RULER). We validate its finite-sample performance using sample splitting and calibrated simulations, and advocate publishing RULER alongside raw rankings so readers can assess both reported performance and the precision of the ordering. RULER also provides guidance for benchmark design that accounts for selection effects and ranking uncertainty.
Source:https://arxiv.org/abs/2611.22227Related posts

I lied to ChatGPT and Gemini and pretended to be an expert
0TechRadar

The overlooked technology that could determine the ROI of enterprise AI
0TechRadar

Microsoft is giving agents their own Cloud PC capabilities to help you with crucial tasks
0TechRadar

The Shure MV360 is an affordable mic for spatial audio capture
0TechRadar

Chief AI Officers are now a common sight in businesses — but new study finds they often don't really have any power
0TechRadar

Nvidia board faction angling for Jensen Huang return as CEO, report claims
0Tom's Hardware