© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI 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. 44437
    Prev
    Next
    INDEX
    Explanations

    references to reports and discussions regarding accountability and oversight in law enforcement

    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
    práv
    -0.14
    ndo
    -0.14
    Interrupt
    -0.14
    ubes
    -0.14
    ÏĢλα
    -0.13
    rgyz
    -0.13
    ella
    -0.13
    .↵↵↵↵↵↵↵↵
    -0.13
    orro
    -0.13
    ocaly
    -0.13
    POSITIVE LOGITS
     neither
    0.42
    Neither
    0.34
     Neither
    0.34
     attempts
    0.31
     spokes
    0.29
     Attempts
    0.28
    Attempts
    0.28
     requests
    0.27
     when
    0.26
     nor
    0.26
    Activations Density 0.030%

    No Known Activations