Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Julius Han · Gated SAE for Llama3-8B-Instruct
    3. Llama3-8B-IT
    4. Residual Stream
    5. 25-RES-JH
    6. 23649
    Prev
    Next
    INDEX
    Explanations

    references to geographic locations or cultural contexts

    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
    iral
    -0.16
    Äįku
    -0.15
     Mercer
    -0.14
    úsqueda
    -0.14
    unu
    -0.14
     Stoke
    -0.14
    assen
    -0.14
    NJ
    -0.14
     Ariel
    -0.14
     Bij
    -0.13
    POSITIVE LOGITS
     Leipzig
    0.37
     Dresden
    0.31
     Leeds
    0.31
     Yorkshire
    0.29
     Sax
    0.26
     Buffalo
    0.25
     Rochester
    0.24
    resden
    0.23
     queryInterface
    0.21
     Sachs
    0.21
    Activations Density 0.002%

    No Known Activations