Nikita Balagansky

elephantmipt's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

authored 4 papers about 11 hours ago

Small Vectors, Big Effects: A Mechanistic Study of RL-Induced Reasoning via Steering Vectors

Paper • 2509.06608 • Published Sep 8, 2025

Steering LLM Reasoning Through Bias-Only Adaptation

Paper • 2505.18706 • Published May 24, 2025

Train One Sparse Autoencoder Across Multiple Sparsity Budgets to Preserve Interpretability and Accuracy

Paper • 2505.24473 • Published May 30, 2025

Interpreting and Steering a Text-to-Speech Language Model with Sparse Autoencoders

Paper • 2606.10029 • Published 3 days ago • 11

submitted a paper to Daily Papers 1 day ago

Interpreting and Steering a Text-to-Speech Language Model with Sparse Autoencoders

Paper • 2606.10029 • Published 3 days ago • 11

authored a paper 10 days ago

Trust-Region Behavior Blending for On-Policy Distillation

Paper • 2605.31159 • Published 14 days ago • 66

authored a paper 3 months ago

Next Embedding Prediction Makes World Models Stronger

Paper • 2603.02765 • Published Mar 3 • 20

authored a paper 11 months ago

Teach Old SAEs New Domain Tricks with Boosting

Paper • 2507.12990 • Published Jul 17, 2025 • 12

authored a paper about 1 year ago

Train Sparse Autoencoders Efficiently by Utilizing Features Correlation

Paper • 2505.22255 • Published May 28, 2025 • 24

authored 4 papers over 1 year ago

You Do Not Fully Utilize Transformer's Representation Capacity

Paper • 2502.09245 • Published Feb 13, 2025 • 37

Analyze Feature Flow to Enhance Interpretation and Steering in Language Models

Paper • 2502.03032 • Published Feb 5, 2025 • 60

Classifiers are Better Experts for Controllable Text Generation

Paper • 2205.07276 • Published May 15, 2022

Mechanistic Permutability: Match Features Across Layers

Paper • 2410.07656 • Published Oct 10, 2024 • 20

authored 3 papers about 2 years ago

PALBERT: Teaching ALBERT to Ponder

Paper • 2204.03276 • Published Apr 7, 2022

Learn Your Reference Model for Real Good Alignment

Paper • 2404.09656 • Published Apr 15, 2024 • 91

Weight Squeezing: Reparameterization for Knowledge Transfer and Model Compression

Paper • 2010.06993 • Published Oct 14, 2020

authored a paper over 2 years ago

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

Paper • 2402.10644 • Published Feb 16, 2024 • 81