AI BENCHMARK PROFILE

AvalancheBench

General AIKnowledge & Reasoning

AvalancheBench evaluates enterprise data agents on latent world recovery, scoring analytical understanding of segments, drivers, temporal events, and relationships from generated observations.

Released: 2026-05-22
Readiness: Paper only
Primary field: General AI

Why it matters

The evaluation gap is the need for controlled diagnostics of whether agents recover analytical structure behind enterprise data, with a rubric for partial credit and error propagation analysis.

Motivation

We introduce AvalancheBench, a benchmark for evaluating enterprise data agents through \emph{latent world recovery}.

Primary resources

Paper ↗Hugging Face ↗

Benchmark Radar records only publicly supported details and links back to primary sources for verification.