Atlas

Models

← All models

Human Laws (SLDBench)

SLDBench Authors · 2025-12-07

Human-derived scaling-law baseline published by the SLDBench authors. It represents the established functional form for each of the benchmark's eight tasks and is stored to preserve the official leaderboard's human comparator.

Benchmark scores

BenchmarkScore
SLDBench Task-Mean R²0.5
Loading Atlas data…