© 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. 19813
    Prev
    Next
    INDEX
    Explanations

    evolution and biology

    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
    è°Ł
    -0.33
    olta
    -0.27
    oS
    -0.25
     ado
    -0.24
    /OR
    -0.24
    ä¸Ģå¦Ĥ
    -0.24
    vro
    -0.24
    ADM
    -0.23
    骢
    -0.23
    isco
    -0.23
    POSITIVE LOGITS
    å¤įæĿĤ
    0.41
    å¤įæĿĤçļĦ
    0.38
     complex
    0.34
     complexity
    0.34
     complexities
    0.32
     fins
    0.31
    Complex
    0.31
     Complexity
    0.30
     Complex
    0.30
    è¤ĩ
    0.29
    Activations Density 0.267%

    No Known Activations