Hjalmar Wijk

Researcher at METR; co-author of RE-Bench and the time-horizon paper.

Role: researcher
Currently at: METR (Model Evaluation and Threat Research)
Twitter: twitter.com/hjalmar_wijk
Papers: 2

Cite

Notes

Only stored in your browser.

Attribution

Affiliations & profile: Semantic Scholar

Attribution policy →

2papers·1eval contribs

Authored papers

Measuring AI Ability to Complete Long Tasks

preprint

2025

RE-Bench: Evaluating frontier AI R&D capabilities of language model agents against human experts

arXiv 2024

2024

Eval contributions

HCAST

METR (Model Evaluation and Threat Research)

METR's Human-Calibrated Autonomy Software Tasks - 189 multi-step software tasks calibrated against human time-to-complete, used to measure agent task length capability.

ActiveCode GenerationPlanningTool CallingAgentic

Affiliations

Currently at

METR (Model Evaluation and Threat Research)

researcher · non profit

Frequent co-authors

from 2 papers

Ben West

researcher

2 shared papers

Brian Goodrich

2 shared papers

Elizabeth Barnes (Beth Barnes)

founder

2 shared papers

Joel Becker

researcher

2 shared papers

Katharyn Garcia

2 shared papers

Lawrence Chan

2 shared papers

Megan Kinniment

researcher

2 shared papers

Michael Chen

researcher

2 shared papers

Neev Parikh

2 shared papers

Nikola Jurkovic

2 shared papers