© 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. 32-TRANSCODER-HP
    7. 42506
    Prev
    Next
    INDEX
    Explanations

    for

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     thá»ķ
    -0.30
    hof
    -0.29
     Brend
    -0.27
    kest
    -0.26
    .units
    -0.26
    ä¸Ģåı¥è¯Ŀ
    -0.26
    NECT
    -0.26
    yun
    -0.25
    inois
    -0.24
    æĬĺå°Ħ
    -0.24
    POSITIVE LOGITS
    storm
    0.31
    Storm
    0.29
    Pe
    0.28
     peel
    0.28
     storm
    0.28
    dbo
    0.27
     Peel
    0.27
    isper
    0.27
    éĿ¢
    0.27
    rah
    0.26
    Activations Density 0.000%

    No Known Activations