ScoringBench: A Benchmark for Evaluating Tabular Foundation Models with Proper Scoring Rules

Jonas Landsgesell, Pascal Knoll · 2026

Read ScoringBench: A Benchmark for Evaluating Tabular Foundation Models with Proper Scoring Rules by Jonas Landsgesell, Pascal Knoll (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-29928.txt.

Description

Landsgesell, Knoll, and Wenzel build ScoringBench for judging tabular foundation models. Such models output full predictive distributions, yet benchmarks mostly grade point guesses. Their bench scores models with proper scoring rules like CRPS and interval scores. It covers 101 regression datasets from many domains. Two ranking schemes are offered, one ordinal and one magnitude-aware. Tests show rankings shift with the scoring rule chosen. Strong point scorers can slip on probabilistic measures.

Catalog record

CREATORJonas Landsgesell, Pascal Knoll
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE33590
LICENSEhttp://creativecommons.org/licenses/by-sa/4.0/
RIGHTSCC BY SA
SOURCEarXiv

Source: https://arxiv.org/abs/2603.29928

Browse the full library · About the Tower of Babel