© 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. 31-TRANSCODER-HP
    7. 18786
    Prev
    Next
    INDEX
    Explanations

    discour

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    å°Ķæĸ¯
    -0.30
     Mig
    -0.30
    ko
    -0.28
    kip
    -0.27
    pei
    -0.27
    oping
    -0.26
    QS
    -0.25
    ican
    -0.25
    cope
    -0.25
    NES
    -0.25
    POSITIVE LOGITS
    å¸ĮæľĽ
    0.34
     CHE
    0.27
     eventual
    0.27
    /bower
    0.27
    :border
    0.26
     hope
    0.26
    ä¹Łå¸ĮæľĽ
    0.25
    rogram
    0.25
     dep
    0.24
     dispersion
    0.24
    Activations Density 0.069%

    No Known Activations