Search Results for author:
Found 6 papers, 4 papers with code
Why Has Predicting Downstream Capabilities of Frontier AI Models with Scale Remained Elusive?
We then reveal the mechanism causing this degradation: downstream metrics require comparing the correct choice against a small number of specific incorrect choices, meaning accurately predicting downstream capabilities requires predicting not just how probability mass concentrates on the correct choice with scale, but also how probability mass fluctuates on specific incorrect choices with scale.
Societal Adaptation to Advanced AI
Existing strategies for managing risks from advanced AI systems often focus on affecting what AI systems are developed and how they diffuse.
The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning
To measure these risks of malicious use, government institutions and major AI labs are developing evaluations for hazardous capabilities in LLMs.
Escalation Risks from Language Models in Military and Diplomatic Decision-Making
Governments are increasingly considering integrating autonomous AI agents in high-stakes military and foreign-policy decision-making, especially with the emergence of advanced generative AI models like GPT-4.
SuperHF: Supervised Iterative Learning from Human Feedback
Here, we focus on two prevalent methods used to align these models, Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF).
Welfare Diplomacy: Benchmarking Language Model Cooperation
The growing capabilities and increasingly widespread deployment of AI systems necessitate robust benchmarks for measuring their cooperative capabilities.
