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. 31298
    Prev
    Next
    INDEX
    Explanations

    references to RSS feeds and related services

    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
    tem
    -0.19
    ovu
    -0.18
    ocht
    -0.18
    tae
    -0.18
    emann
    -0.17
    pty
    -0.17
    ت
    -0.17
    ize
    -0.16
    à¥įध
    -0.16
    fty
    -0.16
    POSITIVE LOGITS
    s
    0.45
    ively
    0.32
    ses
    0.32
    (es
    0.32
    sing
    0.31
    sed
    0.29
    ive
    0.27
    achusetts
    0.27
    ional
    0.27
    es
    0.26
    Activations Density 0.039%

    No Known Activations