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. 23163
    Prev
    Next
    INDEX
    Explanations

    instances of the word "again."

    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
    again
    -0.31
     again
    -0.31
     Again
    -0.29
    Again
    -0.28
     lại
    -0.23
     AGAIN
    -0.23
     novamente
    -0.22
    _again
    -0.22
     ëĭ¤ìĭľ
    -0.20
    åıĪ
    -0.20
    POSITIVE LOGITS
    s
    0.33
    ovnÄĽ
    0.27
    Ùĩ
    0.18
    ÏĤ
    0.17
    oplast
    0.17
    ê¸Ī
    0.17
    sand
    0.17
    sburg
    0.16
    iflower
    0.15
    à¸Ĺ
    0.15
    Activations Density 0.034%

    No Known Activations