© 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. 22-TRANSCODER-HP
    7. 97079
    Prev
    Next
    INDEX
    Explanations

    Code and formulas

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    èµ·æĿ¥äºĨ
    -0.31
    /her
    -0.28
    imat
    -0.26
    Otherwise
    -0.26
     otherwise
    -0.26
    èī¯å¿ĥ
    -0.25
    EH
    -0.25
    ça
    -0.25
    _ib
    -0.24
    eller
    -0.24
    POSITIVE LOGITS
    IOD
    0.25
     {}\
    0.25
    è¿ĻçĤ¹
    0.23
    ä¿Ŀ管
    0.23
    edith
    0.23
    (Display
    0.23
    зол
    0.23
    atten
    0.23
     boast
    0.23
     abolition
    0.23
    Activations Density 0.030%

    No Known Activations