Skip to content
Longterm Wiki

Alignment Evaluations

Evaluationactive

Evaluations specifically designed to measure alignment properties: honesty, helpfulness, harmlessness, and value adherence.

Organizations
4
Grants
3
Total Funding
$446K
Cluster: Evaluation
Parent Area: AI Evaluations

Tags

evaluationsalignmentsafety

Grants3

NameRecipientAmountFunderDate
Operating Capital for AI Safety Evaluation InfrastructureChris Canal$400KManifund2025-10-29
4-month grant to conduct deceptive alignment evaluation research and explore control and mitigation strategiesKai Fronsdal$27KLong-Term Future Fund (LTFF)2024-07
1-month pt. stipend for 4 MATS scholars working on autonomous web-browsing LLM agents that can hire humans + safety evalsSumeet Motwani$19KLong-Term Future Fund (LTFF)2024-01

Funding by Funder

FunderGrantsTotal Amount
Manifund1$400K
Long-Term Future Fund (LTFF)2$46K

Sub-Areas1

NameStatusOrgsPapers
Epistemic Virtue EvaluationsTesting AI systems for epistemic honesty, calibration, and intellectual humility.emerging00