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. 23143
    Prev
    Next
    INDEX
    Explanations

    phrases that describe a player's role and attributes in a sports context

    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
     Palestin
    -0.17
    uars
    -0.15
    ":"/
    -0.14
    á»ĵi
    -0.14
     Gravity
    -0.14
     Baldwin
    -0.13
    >tag
    -0.13
     Campo
    -0.13
    ÏĢε
    -0.13
     Cubs
    -0.13
    POSITIVE LOGITS
     ice
    0.28
     puck
    0.26
     NHL
    0.24
     penalty
    0.23
     goals
    0.23
     goal
    0.23
     Ice
    0.22
     hockey
    0.21
    Ice
    0.21
    goal
    0.20
    Activations Density 0.004%

    No Known Activations