© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 25-TRANSCODER-HP
    7. 109186
    Prev
    Next
    INDEX
    Explanations

    Health and business

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Wunused
    -0.26
     dine
    -0.25
    loff
    -0.25
    éĺ²èĮĥ
    -0.24
     bisc
    -0.24
    äºĮæ°§åĮĸ
    -0.24
    /cli
    -0.24
     liá»ĩt
    -0.23
    egov
    -0.23
     collect
    -0.23
    POSITIVE LOGITS
    mó
    0.28
    thing
    0.28
    ained
    0.28
    art
    0.27
     health
    0.27
    ç§ī
    0.27
    mo
    0.27
     HEALTH
    0.26
    em
    0.26
    åģ¥åº·
    0.26
    Activations Density 0.161%

    No Known Activations