© 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. 18816
    Prev
    Next
    INDEX
    Explanations

    sports and competition

    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
    ToDevice
    -0.25
     unfavor
    -0.24
    ETS
    -0.24
    (ix
    -0.23
    iman
    -0.23
    _UT
    -0.23
     disks
    -0.23
    /libs
    -0.23
     Jung
    -0.22
    emplates
    -0.22
    POSITIVE LOGITS
     entering
    1.02
     going
    0.92
     heading
    0.87
     enter
    0.84
    è¿Ľåħ¥
    0.79
     Enter
    0.77
     enters
    0.77
    going
    0.77
    enter
    0.76
    Entering
    0.74
    Activations Density 42.140%

    No Known Activations