Scholar
Kola Ayonrinde
Google Scholar ID: j40ixccAAAAJ
UK AI Safety Institute
Citations & Impact
Citations
73
H-index
4
i10-index
3
Publications
10
Co-authors
12
list available
Publications
Resume (English only)
Academic Achievements
- Jan 1, 2025: Shazeer Typing
- Dec 11, 2024: SAEBench: A Comprehensive Benchmark for Sparse Autoencoders
- Oct 30, 2024: Standard SAEs Might Be Incoherent: A Choosing Problem & A “Concise” Solution
- Aug 23, 2024: MDL-SAEs: Interpretability as Compression
- Feb 11, 2024: Mamba Explained
- Jan 14, 2024: The Impact of Mixtral
- Jan 8, 2024: Descriptive Matrix Operations with Einops
- Nov 3, 2023: Dictionary Learning with Sparse AutoEncoders
- Oct 22, 2023: An Analogy for Understanding Mixture of Expert Models
- Oct 20, 2023: From Sparse To Soft Mixtures of Experts
- Jul 14, 2023: DeepSpeed's Bag of Tricks for Speed & Scale
Research Experience
- Focused on research involving sparse autoencoders, including but not limited to interpretability through compression techniques and understanding mixture of expert models.
Background
- Research Scientist/ML Engineer
