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. 23225
    Prev
    Next
    INDEX
    Explanations

    phrases indicating comparative amounts or magnitudes

    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
     or
    -0.17
     utmost
    -0.15
    /her
    -0.14
    æĪĸ
    -0.14
    gili
    -0.13
    alto
    -0.13
    ard
    -0.13
     !
    -0.13
    -age
    -0.12
     rather
    -0.12
    POSITIVE LOGITS
     sayıda
    0.17
    ÏĦεÏģο
    0.15
     ÙħÛĮÙĦادÛĮ
    0.14
    ByPrimaryKey
    0.13
    imity
    0.13
    _THAN
    0.13
    entifier
    0.13
    utomation
    0.13
    emailer
    0.13
    нообÑĢаз
    0.13
    Activations Density 0.084%

    No Known Activations