© 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. 20-TRANSCODER-HP
    7. 108464
    Prev
    Next
    INDEX
    Explanations

    programming/data analysis

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     vs
    -0.32
    æ·Ģ
    -0.31
    irc
    -0.26
    efa
    -0.26
     Jew
    -0.26
     Fig
    -0.25
     Pier
    -0.25
    ä¸ľè¥¿
    -0.25
    ey
    -0.24
     ey
    -0.24
    POSITIVE LOGITS
    æĢĽ
    0.27
     controlling
    0.26
     Oswald
    0.25
    owl
    0.25
    çĴIJ
    0.25
    :absolute
    0.24
    /legal
    0.24
    åı»
    0.23
    porto
    0.23
    à¹Īà¸Ńà¸ĩ
    0.23
    Activations Density 11.399%

    No Known Activations