Confidence Regulation Neurons in Language Models
arxiv.org · 5,638 words · saved by 1 readers
N/A
Confidence Regulation Neurons in Language Models Alessandro Stolfo∗ Ben Wu∗ Wes Gurnee ETH Zürich University of Sheffield MIT Yonatan Belinkov Xingyi Song Mrinmaya Sachan Neel Nanda Technion University of Sheffield ETH Zürich arXiv:2406.16254v2 [cs.LG] 8…
saved by
related reading
- Sparse Autoencoders Find Highly Interpretable Features in Language Modelsarxiv.org
- Verbalizable Representations Form a Global Workspace in Language Modelstransformer-circuits.pub
- Language models can explain neurons in language modelsopenaipublic.blob.core.windows.net
- [2305.18290] Direct Preference Optimization: Your Language Model is Secretly a Reward Modelarxiv.org
- [2506.01939] Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoningarxiv.org
- Emotion Concepts and their Function in a Large Language Modeltransformer-circuits.pub
- Language Model Circuits Are Sparse in the Neuron Basis | Transluce AItransluce.org
- [2602.05970] Inverse Depth Scaling From Most Layers Being Similararxiv.org
- [2511.08579] Training Language Models to Explain Their Own Computationsarxiv.org
- Softmax Linear Unitstransformer-circuits.pub
- [2605.27288] It's Not Always Sycophancy: Measuring LLM Conformity as a Function of Epistemic Uncertaintyarxiv.org
- A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Modelsarxiv.org