© 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. 142369
    Prev
    Next
    INDEX
    Explanations

    Punctuation and common words

    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
    大大æıIJé«ĺ
    -0.26
     validity
    -0.26
    çİ®
    -0.25
     prop
    -0.25
     zale
    -0.25
     propagation
    -0.25
     protections
    -0.24
    èĬ°
    -0.24
     props
    -0.24
    broken
    -0.23
    POSITIVE LOGITS
     Fior
    0.28
    ennessee
    0.27
    KHR
    0.26
     fint
    0.24
    åĮĿ
    0.24
    isd
    0.24
    ç«ĭæ¡Ī
    0.24
    tfoot
    0.23
    etal
    0.23
    转è¿IJ
    0.23
    Activations Density 15.178%

    No Known Activations