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. 23192
    Prev
    Next
    INDEX
    Explanations

    phrases related to the concept of gaining attention

    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
     NGX
    -0.17
    iras
    -0.16
    nels
    -0.16
    NSBundle
    -0.15
    liament
    -0.15
    ิà¸ĸ
    -0.14
    woods
    -0.14
    окон
    -0.14
    PIP
    -0.14
    ilate
    -0.14
    POSITIVE LOGITS
     attention
    0.48
     buzz
    0.38
    attention
    0.36
     exposure
    0.34
     Attention
    0.34
    Attention
    0.33
     coverage
    0.33
     press
    0.32
     notice
    0.29
     publicity
    0.28
    Activations Density 0.068%

    No Known Activations