© 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. 140794
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    pheres
    -0.27
    ilen
    -0.26
     preferences
    -0.26
    odia
    -0.26
    rios
    -0.25
     preference
    -0.25
    ä¸ĥ个
    -0.25
    cription
    -0.24
    çĽĹ
    -0.24
    enzyme
    -0.24
    POSITIVE LOGITS
     Overnight
    0.26
    æħ¢æħ¢åľ°
    0.25
     increments
    0.25
    .createCell
    0.25
    overn
    0.24
    æīŃæĽ²
    0.24
    åĨ²
    0.24
    ä¸Ģå¤ľ
    0.24
    éĴ¿
    0.24
    yte
    0.23
    Activations Density 5.068%

    No Known Activations

    This feature has no known activations.