© 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. 26183
    Prev
    Next
    INDEX
    Explanations

    sequences involving numerical data and variables

    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
     E
    -0.15
     M
    -0.14
     Cup
    -0.14
    _scheduler
    -0.14
    .managed
    -0.13
     Comm
    -0.13
    %M
    -0.13
     Q
    -0.13
    wild
    -0.13
     H
    -0.13
    POSITIVE LOGITS
    acles
    0.16
    तर
    0.15
    entai
    0.15
    inous
    0.15
     lÃŃ
    0.15
    yla
    0.15
    bersome
    0.14
    utzer
    0.14
    ños
    0.14
    à¹Īà¸Ńส
    0.14
    Activations Density 0.049%

    No Known Activations