© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 20-TRANSCODER-HP
    7. 63337
    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
    Not in Any Lists

    No Comments

    Negative Logits
    è¡Ķæİ¥
    -0.30
    as
    -0.28
     worm
    -0.27
     Latin
    -0.27
     invol
    -0.26
     destination
    -0.26
     sens
    -0.25
    t
    -0.25
    _dest
    -0.25
     def
    -0.24
    POSITIVE LOGITS
    olls
    0.33
    [now
    0.29
    istine
    0.28
    .inner
    0.28
     [["
    0.26
    碴
    0.26
    .present
    0.26
    <small
    0.25
     Variation
    0.25
    oplay
    0.25
    Activations Density 52.534%

    No Known Activations

    This feature has no known activations.