Boris Shaposhnikov's picture

7

Boris Shaposhnikov

borisshapa

·

borisshapa

AI & ML interests

NLP

Recent Activity

upvoted a paper about 1 month ago

Train Sparse Autoencoders Efficiently by Utilizing Features Correlation

updated a model 3 months ago

borisshapa/sft-llama3.1-8b-uch

published a model 3 months ago

borisshapa/sft-llama3.1-8b-uch

View all activity

Organizations

None yet

authored a paper 5 months ago

The Differences Between Direct Alignment Algorithms are a Blur

Paper • 2502.01237 • Published Feb 3 • 115

authored 2 papers about 1 year ago

Learn Your Reference Model for Real Good Alignment

Paper • 2404.09656 • Published Apr 15, 2024 • 88

Linear Transformers with Learnable Kernel Functions are Better In-Context Models

Paper • 2402.10644 • Published Feb 16, 2024 • 82