© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Julius Han · Gated SAE for Llama3-8B-Instruct
    5. Llama3-8B-IT
    7. Residual Stream
    9. 25-RES-JH
    11. 26062
    Prev
    Next
    INDEX
    Explanations

    references to pockets and the act of retrieving items from them

    oai_token-act-pair · gpt-4o-mini
    Triggered 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
    Negative Logits
    thon
    -0.17
    erland
    -0.17
    stro
    -0.16
    ican
    -0.15
    itet
    -0.15
    лÑĥг
    -0.14
    adders
    -0.14
    екаÑĢ
    -0.14
    dum
    -0.14
    Disposition
    -0.14
    POSITIVE LOGITS
     pocket
    0.57
     pockets
    0.52
     Pocket
    0.43
    Pocket
    0.41
    -pocket
    0.41
     wallet
    0.28
    ocket
    0.27
     purse
    0.26
    ockets
    0.24
     Wallet
    0.23
    Activations Density 0.026%

    No Known Activations