© 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. 9-TRANSCODER-HP
    7. 29132
    Prev
    Next
    INDEX
    Explanations

    and

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_9.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.9.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     win
    -0.27
    对ç«ĭ
    -0.26
     agents
    -0.26
     anc
    -0.26
     unseen
    -0.26
    åĪĨå·¥
    -0.25
    æĿ²
    -0.25
     rig
    -0.25
     article
    -0.25
    oks
    -0.25
    POSITIVE LOGITS
    第ä¸Ģ个
    0.31
    çķĻåľ¨
    0.29
    çļĦ第ä¸Ģ个
    0.27
    ender
    0.27
     engagements
    0.26
     accompagn
    0.26
     Twice
    0.25
    ữ
    0.25
    keep
    0.25
    ÙĦÙħ
    0.24
    Activations Density 1.091%

    No Known Activations