Skip to content
conftrace_

mlp layer

5 papers

Explore in graph

Co-occurring keywords · papers mentioning both

mechanistic interpretability (2) mlp block (1) modular addition (1) attention layer (1) feature circuit (1) reasoning analysis (1) mechanistic analysis (1) gradient sparsity (1) neural network (1) large language model (1)

Papers

Dissecting Persona-Driven Reasoning in Language Models via Activation Patching EMNLP 2025 Learning Neural Networks with Sparse Activations COLT 2024 Pre-trained Large Language Models Use Fourier Features to Compute Addition NeurIPS 2024 SparseGrad: A Selective Method for Efficient Fine-tuning of MLP Layers EMNLP 2024 Transcoders find interpretable LLM feature circuits NeurIPS 2024