cd/entity/SciBench· home› entities› SciBench
grep -l @scibench /news/*.json | wc -l → 1

SciBench

mentions 1 type Organization feed RSS

// recent coverage 1 mentions

04:00
2026-09-30
arxiv.org
large-language-models

Binarization Flattens the Score Space

A new arXiv paper (2609.35797v1) finds that collapsing large language model judge rewards to pass/fail {0, 1} hides proportional changes in how well a response meets each criterion, and recommends kee…

// co-occurs with top 3 entities
// topics top 4 topics