© 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. 33-TRANSCODER-HP
    7. 54236
    Prev
    Next
    INDEX
    Explanations

    Foreign language fragments

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    弧
    -0.27
    æŶåĢĻ
    -0.27
     Diá»ħn
    -0.26
    arium
    -0.26
    ARC
    -0.25
    俾
    -0.24
    å·¡
    -0.24
    ios
    -0.24
    Arc
    -0.24
    encers
    -0.24
    POSITIVE LOGITS
    aget
    0.28
     CA
    0.27
    è¿Ľ
    0.26
     chÃŃ
    0.26
    ,tr
    0.24
    «
    0.24
     jc
    0.24
    éĢ²
    0.24
     conscience
    0.23
    éĽĨåĽ¢èĤ¡ä»½
    0.23
    Activations Density 0.007%

    No Known Activations