© 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. 19443
    Prev
    Next
    INDEX
    Explanations

    power

    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
     CALLBACK
    -0.27
    shed
    -0.26
    åħ«åŃĹ
    -0.24
     shed
    -0.24
    arn
    -0.23
     Bach
    -0.23
    hem
    -0.23
    éĻIJ度
    -0.23
     Rosen
    -0.22
    _CRITICAL
    -0.22
    POSITIVE LOGITS
     power
    0.69
     powers
    0.65
    åĬĽéĩı
    0.60
     poder
    0.59
    å¹Ĥ
    0.59
    -p
    0.58
    power
    0.57
    æĿĥåĬĽ
    0.57
     pow
    0.56
    powers
    0.56
    Activations Density 18.319%

    No Known Activations