© 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. 19169
    Prev
    Next
    INDEX
    Explanations

    mentions locations

    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
     SY
    -0.27
    稳
    -0.27
    lo
    -0.27
    Sym
    -0.26
    ANK
    -0.26
     Sym
    -0.26
     bedding
    -0.26
    enko
    -0.25
    sym
    -0.25
    身åŃIJ
    -0.24
    POSITIVE LOGITS
     Kimberly
    0.33
     pearls
    0.30
    ico
    0.28
    alice
    0.28
    /ar
    0.26
    estone
    0.26
    RAL
    0.26
    Kim
    0.25
     Kimber
    0.25
    icons
    0.25
    Activations Density 7.356%

    No Known Activations