© 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. 19-TRANSCODER-HP
    7. 8722
    Prev
    Next
    INDEX
    Explanations

    code and links

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_19.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.19.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    æ¯Ķäºļ
    -0.28
    inations
    -0.27
    ç³Ĭ
    -0.26
    jaw
    -0.26
    eses
    -0.25
     interpretations
    -0.24
    isy
    -0.24
    duc
    -0.23
     Giuliani
    -0.23
    icity
    -0.23
    POSITIVE LOGITS
     Liquid
    0.24
    iless
    0.24
    迨
    0.23
    ÏĪ
    0.23
    ernel
    0.23
    _nc
    0.23
    .Attributes
    0.23
    ector
    0.23
    spin
    0.23
    _increment
    0.23
    Activations Density 0.013%

    No Known Activations