© 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. 30-TRANSCODER-HP
    7. 96435
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    alist
    -0.28
    æĮij
    -0.27
     parch
    -0.26
    _pe
    -0.26
    reon
    -0.25
    æł©
    -0.25
    éĺ²èħIJ
    -0.25
     alist
    -0.25
    æĵį
    -0.25
    ipe
    -0.24
    POSITIVE LOGITS
     automatically
    0.29
     Cups
    0.27
    æ°Ĺãģ«
    0.26
    едÑĮ
    0.26
     Automatically
    0.26
    èİ«
    0.25
    æī«ä¸Ģ
    0.25
    male
    0.25
     Bras
    0.24
     Dw
    0.24
    Activations Density 0.882%

    No Known Activations

    This feature has no known activations.