Running 4 CorrSteer: Correlation-Based Steering of Language Models via Sparse Autoencoders 🧭 4 Steer language model output by clicking visual layers
CorrSteer: Steering Improves Task Performance and Safety in LLMs through Correlation-based Sparse Autoencoder Feature Selection Paper • 2508.12535 • Published Aug 18, 2025 • 2
Running 80 Bringing paper to life: A modern template for scientific writing 📝 80 Explore an interactive galaxy visualization of scientific article
FaithfulSAE: Towards Capturing Faithful Features with Sparse Autoencoders without External Dataset Dependencies Paper • 2506.17673 • Published Jun 21, 2025 • 7