Papers should add a section benchmarking frontier models, argues VIGA thread on broken peer review

FinanceYF5 · x · 2026-09-18

Continuing the Michael Black / VIGA thread, the author argues academic peer review remains fixated on 'technical novelty,' pushing researchers to tweak architectures rather than move the field forward.

Proposal: papers should add a new section, like 'Related Work,' that evaluates current frontier models on the task — failing to try Astra or successor models means the survey is incomplete.

Related event: Scholars argue research should first stress-test existing LLMs(2 posts)→

Original post →

More from Research

Research channel →