© 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. 96604
    Prev
    Next
    INDEX
    Explanations

    surface

    np_max-act-logits · gemini-2.0-flash
    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
     exists
    -0.27
     skate
    -0.26
    åŃĺåľ¨
    -0.26
     existence
    -0.26
     Relay
    -0.26
     happened
    -0.26
     presence
    -0.25
    åŃĺåľ¨çļĦ
    -0.25
     exist
    -0.24
    born
    -0.24
    POSITIVE LOGITS
     surface
    0.64
    surface
    0.61
     surfaces
    0.61
     Surface
    0.58
    Surface
    0.57
    表éĿ¢
    0.51
    _surface
    0.50
    urface
    0.48
    .surface
    0.48
     surf
    0.47
    Activations Density 0.102%

    No Known Activations