© 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. 32-TRANSCODER-HP
    7. 43618
    Prev
    Next
    INDEX
    Explanations

    hath

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     tst
    -0.27
    æĬ½è±¡
    -0.27
    yaw
    -0.27
    çľĭä»ĸ
    -0.26
    åı«æĪij
    -0.26
    æŀķ
    -0.25
    æģķ
    -0.25
     cdecl
    -0.24
    åħ¥åĽŃ
    -0.24
     Wyn
    -0.24
    POSITIVE LOGITS
     share
    0.33
    ady
    0.29
    SHARE
    0.29
     Share
    0.27
     power
    0.27
    Share
    0.27
     place
    0.27
    geme
    0.27
    inking
    0.26
    uding
    0.26
    Activations Density 0.001%

    No Known Activations