© 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. 130617
    Prev
    Next
    INDEX
    Explanations

    Update or Edit

    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
    åIJ¬è¯´è¿ĩ
    -0.28
    enton
    -0.27
    /DD
    -0.27
    鬲
    -0.25
    INA
    -0.25
    çļ¿
    -0.25
    æĸĽ
    -0.24
    axe
    -0.24
    ights
    -0.24
    å®ļåIJij
    -0.24
    POSITIVE LOGITS
    èĨĬ
    0.28
    _drv
    0.27
     turnaround
    0.25
     modification
    0.25
    Ñĥв
    0.24
    Above
    0.24
    -looking
    0.24
    '][]
    0.23
    ÑĮ
    0.23
    ноÑģÑĤ
    0.23
    Activations Density 1.896%

    No Known Activations