© 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. 25194
    Prev
    Next
    INDEX
    Explanations

    symbols or notations related to mathematical or programming expressions

    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
    most
    -0.19
    /latest
    -0.16
    úsqueda
    -0.15
    egers
    -0.15
    adder
    -0.15
    usan
    -0.15
    à¹Ĩ
    -0.15
    UCH
    -0.15
    /original
    -0.15
    evin
    -0.15
    POSITIVE LOGITS
    ãĤĪãģĨãģ«
    0.15
    è°±
    0.15
    ILLISE
    0.14
    ãģ°
    0.14
    ÑĪÑĤов
    0.14
    yard
    0.14
    ãĤ¡
    0.13
    ëŁī
    0.13
    ส
    0.13
     tiener
    0.13
    Activations Density 0.003%

    No Known Activations