11,000 papers. 10 worth reading. How do you find them?

PapersWithELO is not an AI reviewer. It is an open, incremental attention ranking protocol — it tells you what you can safely skip.

90%
of the bottom 25% is rejected
base reject rate: 68.9%
$4.74
per 100 papers, two judges
4,280 pairwise observations
0.75
AUC, fused θ vs. decisions
model-agnostic: mimo × deepseek fusion

A filter, not a chooser

Distribution of ground-truth decisions across fused percentile, ICLR 2024 (10% sample, seed 42, n = 730).

We are good at filtering low-quality work, not picking the best. The top decile is still 69% accepted — as it should be; science is noisy.

How it works

1. Batch

6 papers per batch. The LLM never sees an absolute scale.

2. Rank

The judge only outputs an ordering. Ranking is the only output we trust.

3. Fit

Bayesian Plackett–Luce with MAP + Laplace uncertainty: every paper gets θ ± σ.

4. Fuse

Two judges (mimo, deepseek), z-scored and fused. New entries rank against fixed incumbents.

LLMs are too kind for absolute scores — everything is “above average, 7/10”. We never ask for scores. We force comparisons, and fit a Bayesian model to the rankings.

Three independent tournaments

Each year is ranked separately — θ is only comparable within a year; percentile is comparable across years.