© 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. 31-TRANSCODER-HP
    7. 19789
    Prev
    Next
    INDEX
    Explanations

    by

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     by
    -0.41
    by
    -0.38
    (by
    -0.35
     oleh
    -0.32
    By
    -0.31
    /by
    -0.31
    BY
    -0.31
     By
    -0.30
    .by
    -0.30
    _by
    -0.30
    POSITIVE LOGITS
    ä¸ĢèĦ¸
    0.28
    å¿ĥ
    0.27
     then
    0.27
    ä»ĬæĻļ
    0.27
     heart
    0.27
    then
    0.27
    ä»Ĭ天
    0.27
    Then
    0.26
     tossing
    0.26
     THEN
    0.26
    Activations Density 0.250%

    No Known Activations