© 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. 44345
    Prev
    Next
    INDEX
    Explanations

    mathematical terminology and concepts related to the Mandelbrot set and its properties

    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
     predecess
    -0.14
    preced
    -0.14
    Tho
    -0.14
    ãĥ¬ãĤ¹
    -0.13
    ãĥ³ãĥĸ
    -0.13
     nouve
    -0.13
    ез
    -0.12
    ofday
    -0.12
    /release
    -0.12
     doz
    -0.12
    POSITIVE LOGITS
    acon
    0.13
     iParam
    0.13
     we
    0.12
    orum
    0.12
     therefore
    0.12
    ittel
    0.12
    isin
    0.12
    PÅĻi
    0.12
    uffman
    0.12
    Ä
    0.12
    Activations Density 0.135%

    No Known Activations