© 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. 20-TRANSCODER-HP
    7. 108224
    Prev
    Next
    INDEX
    Explanations

    Code/text separators

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     addAction
    -0.26
     onBind
    -0.26
    adc
    -0.25
     koÅĦ
    -0.25
    wand
    -0.24
    æŁ¥çľĭåħ¨æĸĩ
    -0.24
     endIndex
    -0.24
    _PI
    -0.24
    .pi
    -0.24
    èĹł
    -0.24
    POSITIVE LOGITS
    bel
    0.30
    uration
    0.28
    åħ¬å¸ĥçļĦ
    0.26
    out
    0.26
     Wheel
    0.26
    è¿ĻæŶåĢĻ
    0.25
    ertainment
    0.25
    Bel
    0.24
    Western
    0.24
    ouro
    0.24
    Activations Density 7.403%

    No Known Activations