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. 23332
    Prev
    Next
    INDEX
    Explanations

    references to education and related policies

    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
     Bakery
    -0.16
    anela
    -0.15
     consum
    -0.14
    aggi
    -0.14
    acro
    -0.14
     éĩ
    -0.14
    elah
    -0.14
     reward
    -0.14
     humanitarian
    -0.14
    èĩ¨
    -0.13
    POSITIVE LOGITS
     universal
    0.42
    universal
    0.41
    Universal
    0.37
     free
    0.37
     Universal
    0.36
    åħįè´¹
    0.34
     univers
    0.33
    free
    0.33
     FREE
    0.31
     UNIVERS
    0.28
    Activations Density 0.061%

    No Known Activations