© 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. 18789
    Prev
    Next
    INDEX
    Explanations

    Code and technical language

    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
    uze
    -0.28
     Vo
    -0.28
    沸
    -0.26
     sponge
    -0.26
    èĮģ
    -0.26
    VO
    -0.26
    -growing
    -0.26
    éķĽ
    -0.26
    Vo
    -0.26
    eren
    -0.26
    POSITIVE LOGITS
    ars
    0.29
    iod
    0.27
     Elliott
    0.27
     cps
    0.26
    ennes
    0.26
    .ls
    0.26
    è°Ĵ
    0.26
    γ
    0.25
    arr
    0.25
    éĺ¿æĭī伯
    0.25
    Activations Density 0.393%

    No Known Activations