© 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. 130032
    Prev
    Next
    INDEX
    Explanations

    osphere

    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
    æĹłåģ¿
    -0.27
    _except
    -0.27
    urge
    -0.26
    çļĦä¸Ģ项
    -0.25
    ,Yes
    -0.25
    åijĬ
    -0.25
    ;y
    -0.24
    while
    -0.24
    Bond
    -0.23
    krä
    -0.23
    POSITIVE LOGITS
     things
    0.29
     decent
    0.27
     fairly
    0.26
     promising
    0.26
     viable
    0.25
    åIJĪ并
    0.25
     proper
    0.24
    -compatible
    0.24
    ä½ľèĢħ
    0.23
     relatively
    0.23
    Activations Density 2.915%

    No Known Activations