© 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. 35-TRANSCODER-HP
    7. 38759
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    ION
    -0.28
    è¤IJ
    -0.28
     cannon
    -0.28
    BB
    -0.27
     Bolshevik
    -0.25
     propulsion
    -0.25
    ursive
    -0.24
    ä¸Ńåįİ人æ°ij
    -0.24
    è¿ĺ没æľī
    -0.24
    ion
    -0.24
    POSITIVE LOGITS
    -products
    0.27
    -pencil
    0.27
    -Day
    0.26
    æĻ®éĢļ人
    0.26
    -directed
    0.26
    krä
    0.25
    celed
    0.25
     scrap
    0.25
    é«ĺå³°
    0.25
    -mort
    0.24
    Activations Density 0.001%

    No Known Activations

    This feature has no known activations.