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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI 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. 55101
    Prev
    Next
    INDEX
    Explanations

    references to specific geographical regions and their associated political or military contexts

    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
     Memphis
    -0.20
     Puerto
    -0.19
     Chennai
    -0.18
     Egyptian
    -0.17
    Egypt
    -0.17
     Mississippi
    -0.17
     Singapore
    -0.17
    -Israel
    -0.16
    Singapore
    -0.16
     Lagos
    -0.16
    POSITIVE LOGITS
     Afghanistan
    0.59
     Afghan
    0.58
     Kabul
    0.49
     Taliban
    0.46
    Af
    0.43
     Af
    0.40
     af
    0.36
    .af
    0.33
     Pas
    0.30
    #af
    0.29
    Activations Density 0.015%

    No Known Activations