Skip to content
Longterm Wiki
Index
Grant·AnvK38WznP·Record·Profile

Grant: Straumli — LLM Cyberoffense Benchmark (Coefficient Giving → Straumli)

Verdictconfirmed95%
1 check · 4/9/2026

Deterministic match: grantee, amount, date matched in source snapshot (2714 rows)

Our claim

entire record
Name
Straumli — LLM Cyberoffense Benchmark
Amount
$400,000
Currency
USD
Date
February 2024
Notes
[Navigating Transformative AI] Open Philanthropy recommended a grant of $400,000 to Straumli to support a project led by Paul Bricman to develop a benchmark for the cyberoffense capabilities of Large Language Model (LLM) agents. This grant was funded via a request for proposals expand[Navigating Transformative AI] Open Philanthropy recommended a grant of $400,000 to Straumli to support a project led by Paul Bricman to develop a benchmark for the cyberoffense capabilities of Large Language Model (LLM) agents. This grant was funded via a request for proposals for projects benchmarking LLM agents on consequential real-world tasks and falls within our focus area of potential risks from advanced artificial intelligence.

Source evidence

1 src · 1 check
confirmed95%deterministic-row-match · 4/9/2026
Name
Straumli — LLM Cyberoffense Benchmark
Grantee
Straumli
Focus Area
Navigating Transformative AI
Amount
$400,000.00
Date
February 2024
Description
Open Philanthropy recommend

NoteDeterministic match: grantee, amount, date matched in source snapshot (2714 rows)

Case № AnvK38WznPFiled 4/9/2026Confidence 95%
Source Check: Grant: Straumli — LLM Cyberoffense Benchmark (Coefficient Giving -> Straumli) | Longterm Wiki