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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 28113
    Prev
    Next
    INDEX
    Explanations

    key subjects and their actions in a narrative context

    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
     bottom
    -0.28
    bottom
    -0.22
     Bottom
    -0.21
     top
    -0.20
     bottoms
    -0.20
    -bottom
    -0.19
    Bottom
    -0.17
    _bottom
    -0.17
    .bottom
    -0.17
    top
    -0.16
    POSITIVE LOGITS
     escape
    0.78
     Escape
    0.67
     escapes
    0.64
     escaping
    0.60
     escaped
    0.59
    Escape
    0.59
     esc
    0.59
     Esc
    0.58
    escape
    0.57
     ESC
    0.52
    Activations Density 0.032%

    No Known Activations