© 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. 45048
    Prev
    Next
    INDEX
    Explanations

    phrases related to law enforcement and legal action

    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
    thing
    -0.14
     theirs
    -0.14
     BOTH
    -0.13
    ìŀ¡ëĭ´
    -0.13
    олом
    -0.12
    theless
    -0.12
     strugg
    -0.12
    Ïīμά
    -0.12
    .twig
    -0.12
    ÑĢава
    -0.12
    POSITIVE LOGITS
     various
    0.41
    ses
    0.34
     few
    0.26
     numerous
    0.26
     Various
    0.25
     several
    0.24
    Various
    0.23
     many
    0.23
     aforementioned
    0.23
     handful
    0.23
    Activations Density 0.615%

    No Known Activations