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. 23442
    Prev
    Next
    INDEX
    Explanations

    references to social interactions and responses in public spaces

    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
    Guest
    -0.19
     Guest
    -0.18
     guest
    -0.17
     bedrooms
    -0.15
     barn
    -0.14
     farmhouse
    -0.14
     Guests
    -0.14
    guest
    -0.14
    ientes
    -0.14
     perfor
    -0.14
    POSITIVE LOGITS
     street
    0.39
     passer
    0.36
     pass
    0.35
     passing
    0.33
     pedestrians
    0.33
     sidewalk
    0.31
     pedestrian
    0.31
    street
    0.30
     walking
    0.30
     Street
    0.30
    Activations Density 0.048%

    No Known Activations