© 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. 19529
    Prev
    Next
    INDEX
    Explanations

    family relationships

    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
     maiden
    -0.29
    èİĵ
    -0.28
    igram
    -0.26
     stitches
    -0.26
     Rover
    -0.26
    lef
    -0.25
     clearing
    -0.25
    ç³»
    -0.25
    patch
    -0.25
     cr
    -0.24
    POSITIVE LOGITS
     sibling
    0.27
    eso
    0.27
    æѤæĸĩ
    0.27
    allee
    0.25
     RVA
    0.25
     prospective
    0.24
     parent
    0.24
    йÑĤ
    0.24
    asti
    0.24
    家éķ¿
    0.24
    Activations Density 0.045%

    No Known Activations