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. 31405
    Prev
    Next
    INDEX
    Explanations

    expressions of progression and anticipation

    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
    ModifiedDate
    -0.17
     ******************************************************************************↵
    -0.15
    Ïģε
    -0.14
    toLocale
    -0.14
     hypoth
    -0.14
    -IS
    -0.14
    .gt
    -0.14
    arme
    -0.13
    .nt
    -0.13
    .e
    -0.13
    POSITIVE LOGITS
     barely
    0.43
     scratches
    0.34
     scratched
    0.34
     scratching
    0.34
     scratch
    0.33
     merely
    0.31
     apenas
    0.29
     tip
    0.28
     Scratch
    0.27
    scratch
    0.27
    Activations Density 0.019%

    No Known Activations