default author photo

Sunishchal Dev

RAND Corporation

SCHOLARLY PAPERS

1

DOWNLOADS

74

TOTAL CITATIONS

0

Scholarly Papers (1)

1.

Recommendations and Reporting Checklist for Rigorous & Transparent Human Baselines in Model Evaluations

A version of this paper has been accepted to ICML 2025 as a position paper (spotlight), with the title: "Position: Human Baselines in Model Evaluations Need Rigor and Transparency (With Recommendations & Reporting Checklist).", RAND Working Paper Series Forthcoming
Number of pages: 69 Posted: 22 Jun 2025
RAND Corporation, RAND Corporation, RAND Corporation, RAND Corporation, Stanford University, Harvard University, Harvard University, Max Planck School of Cognition and Harvard University
Downloads 74 (852,604)

Abstract:

Loading...

Human Baseline, Human Performance Baseline, AI Evaluation, ML Evaluation, Evaluation Methodology, Science of Evaluation, Science of Evaluations, AI Evaluations, LLM Evaluation, LLM Evaluations, ML Evaluations, Benchmarks