AI-Driven Language Model Benchmarking @DARPAtv
AI-Driven Language Model Benchmarking  @DARPAtv
Uploaded October 2026 | Updated November 2026, 21 hours ago
Dr. Percy Liang (Stanford) discusses his team’s approach to language model evaluation under the AIQ program. They use AI models to automatically cover the test input space and find benchmarks or other metrics that transfer smoothly across scales. Both of these aims inform the third aim of safety evaluation.

• The views, opinions and/or findings expressed are those of the presenter and should not be interpreted as representing the official views or policies of the Department of Defense or the U.S. Government.

This content was created prior to Executive Order 14434, directing the Government's use of the term Super Intelligence.
AI-Driven Language Model BenchmarkingDARPA Venture Horizons - MTO Spark TankAmazons Formal Methods Journey - Michael HicksClosing Remarks - Kathleen FisherPlaying with Heavy Hitters: A Look Inside DARPAs CultureDARPA Triage Challenge: Integrating Sensor Data for Life-Saving InsightsFrom Data Security to Discovery: How ARPA-H is Using AI to Transform Health Care in AmericaQuantum Benchmarking Initiative OverviewDARPA Lift Challenge - Aug 5 - Main BroadcastNo Manning Required Ship (NOMARS) ChristeningTeam All You Need IS A Fuzzing Brain - DARPA AI Cyber Challenge“Robo Duck” Architecture
DARPAtv |

AI-Driven Language Model Benchmarking

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER