© 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. 140568
    Prev
    Next
    INDEX
    Explanations

    code

    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
    Negative Logits
     coincidence
    -0.30
     coinc
    -0.27
    å¼ĢæĿ¥
    -0.26
     commission
    -0.26
    isma
    -0.26
    ä¸ij
    -0.25
    remember
    -0.25
     energ
    -0.24
    åľ¨ä¸ŃåĽ½
    -0.24
    çļĦäºĭå®ŀ
    -0.24
    POSITIVE LOGITS
     dipping
    0.27
     tempered
    0.27
     somehow
    0.27
     eup
    0.26
     Thor
    0.26
    epy
    0.26
    çĤ¹æ»´
    0.26
     Edgar
    0.25
    uggy
    0.25
    Thor
    0.25
    Activations Density 0.005%

    No Known Activations