© 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 EvalsExportsinterp-engineNEWAPI 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. 21533
    Prev
    Next
    INDEX
    Explanations

    technical terms related to log file management and compression

    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
    .codes
    -0.14
     Fur
    -0.13
    acas
    -0.13
     Dump
    -0.13
    $__
    -0.13
    ewan
    -0.13
    fullscreen
    -0.13
    Mess
    -0.13
    ipsis
    -0.13
    æĬ
    -0.13
    POSITIVE LOGITS
     file
    0.22
     inode
    0.21
     touch
    0.20
    /stat
    0.19
    touch
    0.19
     directory
    0.19
     dirent
    0.19
     contents
    0.18
    stat
    0.18
     files
    0.18
    Activations Density 0.021%

    No Known Activations