© 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. 18559
    Prev
    Next
    INDEX
    Explanations

    of

    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
    unit
    -0.42
     unit
    -0.41
    åįķä½į
    -0.39
     units
    -0.35
    Unit
    -0.33
    åįķåħĥ
    -0.33
     unidad
    -0.33
    _Unit
    -0.32
     unidades
    -0.32
    _UNIT
    -0.31
    POSITIVE LOGITS
     ObjectMapper
    0.25
    пÑĢов
    0.24
    èIJ½
    0.24
    liers
    0.24
     disposition
    0.23
    åıĺå¾ĹæĽ´
    0.23
    åłķ
    0.23
    ä¿¡æģ¯æľįåĬ¡
    0.23
     IMapper
    0.23
    itory
    0.23
    Activations Density 0.126%

    No Known Activations