© 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. 63669
    Prev
    Next
    INDEX
    Explanations

    all

    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
    edom
    -0.26
    åIJijä¸ĭ
    -0.26
    åıĭ们
    -0.26
    åĪģ
    -0.25
    icts
    -0.24
    ',//
    -0.24
    åĴĮæĶ¯æĮģ
    -0.24
    :NS
    -0.24
    .sendStatus
    -0.24
    iset
    -0.24
    POSITIVE LOGITS
    (Max
    0.28
    æµĨ
    0.26
    éĸ
    0.25
    æĹģè¾¹çļĦ
    0.25
     Slut
    0.24
    riter
    0.23
    .ap
    0.23
    第ä¸ĢçĻ¾
    0.23
    apot
    0.23
    nung
    0.23
    Activations Density 8.660%

    No Known Activations