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. 23536
    Prev
    Next
    INDEX
    Explanations

    recommendations for a healthy lifestyle and its components

    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
    à¹īà¸ĩ
    -0.16
    edium
    -0.14
    _ATOM
    -0.14
     пÑĢоз
    -0.13
    384
    -0.13
    eventId
    -0.13
    عÙħ
    -0.13
    ymbol
    -0.13
     verst
    -0.13
    жд
    -0.13
    POSITIVE LOGITS
     exercise
    0.42
     Exercise
    0.36
     regular
    0.32
    exercise
    0.32
     exercises
    0.32
    Exercise
    0.31
     exercising
    0.30
     diet
    0.28
     exercised
    0.27
    regular
    0.27
    Activations Density 0.071%

    No Known Activations