Blog.
May 12, 2024
The Benchmark Trap: Why LLM Metrics Mislead and Evals Enlighten
The multitude of issues with LLM benchmarks, and the problem of documenting the complexities to accurately test LLMs, and using evals as alternatives
Apr 16, 2024
Behind the Curve: The Slow March of AI into Classrooms
The discrepancy between the potential impact of AI, its current limited use in the educational sphere, and the reasons behind it
Feb 2, 2024
Artificial Misdirection: Misalignment of LLMs
The problems with AI alignment, and how instrumental misalignment threatens to worsen it.
