© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 41675
    Prev
    Next
    INDEX
    Explanations

    phrases related to health issues and their impact on individuals and society

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Juliushanhanhan/llama-3-8b-it-res/blocks.25.hook_resid_post
    Features
    65,536
    Data Type
    float32
    Hook Name
    blocks.25.hook_resid_post
    Hook Layer
    25
    Architecture
    gated
    Context Size
    1,024
    Dataset
    Juliushanhanhan/openwebtext-1b-llama3-tokenized-cxt-1024
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     mÃŃn
    -0.17
    <Props
    -0.15
    arer
    -0.15
    idak
    -0.14
     øns
    -0.14
    andes
    -0.13
    -complete
    -0.13
    wich
    -0.13
    sdk
    -0.13
     ank
    -0.13
    POSITIVE LOGITS
     avoid
    0.51
     prevent
    0.50
    prevent
    0.36
     Avoid
    0.35
    avoid
    0.35
    Avoid
    0.32
     unnecessary
    0.30
     avoided
    0.29
     deaths
    0.28
     Prevent
    0.27
    Activations Density 0.086%

    No Known Activations