© 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. 13846
    Prev
    Next
    INDEX
    Explanations

    content that discusses tactics and impacts related to terrorism and psychological effects

    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
    ipur
    -0.16
    Leaks
    -0.15
    Ĺı
    -0.15
     defaultManager
    -0.14
    ragen
    -0.14
    ftype
    -0.14
    enaire
    -0.14
     trục
    -0.14
    á»ı
    -0.14
    olv
    -0.13
    POSITIVE LOGITS
     fear
    1.08
     Fear
    1.04
    Fear
    1.01
     fears
    0.79
     fearful
    0.73
     feared
    0.73
     afraid
    0.71
    æģIJ
    0.66
     fearing
    0.65
     terror
    0.65
    Activations Density 0.020%

    No Known Activations