NumericJev: numerical decoding through structured range choices

A training-free algorithm that recovers numerical outputs by recursively narrowing a range through a multiway tree of Jev-like choices. The paper preserves the original question in context and studies how range selection differs from direct candidate selection.

Notes

On the authors' 100-value-grid arithmetic experiment, a ten-way tree takes two decision rounds and reaches 1.84% range-normalized mean absolute error versus 5.18% for direct choice. Code is released with the preprint.

source-backed — Public repository, docs or live artifact. About this label

Destinations

Filed under

• #research
• #structured-output
• #benchmarks
• #decision-making

In collections

• Start here: official and explainers →
• Benchmarks & independent tests →