© 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. Qwen3-4B
    5. 20-TRANSCODER-HP
    7. 64101
    Prev
    Next
    INDEX
    Explanations

    computer code

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    kl
    -0.28
    iras
    -0.27
    éĢ»è¾ij
    -0.26
    oga
    -0.25
    ))->
    -0.24
    latable
    -0.24
    ÃĤN
    -0.24
    ortho
    -0.24
     Tracy
    -0.24
    kid
    -0.24
    POSITIVE LOGITS
    apos
    0.27
    à¹īม
    0.26
    crease
    0.25
    uzzi
    0.25
    ewire
    0.24
    认
    0.24
    ç§Ģ
    0.23
     creeping
    0.23
    ç¥ļ
    0.23
     prematurely
    0.23
    Activations Density 84.425%

    No Known Activations