© 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. 12901
    Prev
    Next
    INDEX
    Explanations

    references to specific dates and meetings related to Robert Mugabe

    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
    ombie
    -0.14
    ÑĢÑĥ
    -0.14
    optic
    -0.14
    ocoa
    -0.14
    zenia
    -0.14
    ÑıÑģ
    -0.13
    _Impl
    -0.13
     Evel
    -0.13
    adoo
    -0.13
    ZN
    -0.13
    POSITIVE LOGITS
     hide
    1.13
     Hide
    1.09
    Hide
    0.99
     hiding
    0.98
     hides
    0.96
    hide
    0.95
     hidden
    0.93
     hid
    0.92
     Hidden
    0.88
    -hide
    0.84
    Activations Density 0.006%

    No Known Activations