cd/entity/Relay-Bench· home› entities› Relay-Bench
grep -l @relay-bench /news/*.json | wc -l → 1

Relay-Bench

mentions 1 type Organization feed RSS

// recent coverage 1 mentions

04:00
2026-07-22
machinebrief.com
large-language-models

Relay-Bench: Evaluating LLMs on Multi-Domain Reasoning Chains

A new benchmark called Relay-Bench, introduced in arXiv:2607.18438v1, measures large language models' ability to complete multi-domain reasoning chains in a single prompt, with the leading model GPT-5…

// co-occurs with top 2 entities
GPT-5.5 (xHigh) 1 arXiv 1
// topics top 2 topics
large language models 1 ai research 1