© 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. 22-TRANSCODER-HP
    7. 112047
    Prev
    Next
    INDEX
    Explanations

    Project

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åĿ¯
    -0.29
    ละ
    -0.26
    arked
    -0.26
    ingham
    -0.26
    éĴ¦
    -0.26
    inia
    -0.25
    ognito
    -0.25
    ä¿ĥ
    -0.24
    æĪIJåIJį
    -0.24
    sticks
    -0.24
    POSITIVE LOGITS
    iro
    0.27
    ICO
    0.27
     way
    0.25
    esimal
    0.25
    lse
    0.25
    duto
    0.24
    zie
    0.24
    æľīçļĦ
    0.24
    åĴĮæĶ¯æĮģ
    0.24
    ç½ijç«Ļ建设
    0.24
    Activations Density 0.000%

    No Known Activations