© 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. 53339
    Prev
    Next
    INDEX
    Explanations

    Short foreign words

    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
    stin
    -0.32
     yuk
    -0.31
    st
    -0.30
    cl
    -0.30
     st
    -0.29
    ä½ľèĢħæīĢæľī
    -0.28
    ward
    -0.28
    ese
    -0.27
    wards
    -0.27
     needles
    -0.27
    POSITIVE LOGITS
    łģ
    0.26
     Mol
    0.26
    ÑĦеÑĢ
    0.25
    åĨ¶
    0.25
     moth
    0.24
    海峡
    0.24
    鼷æĸ¯
    0.24
    pha
    0.24
     trÆ°á»Łng
    0.23
     Nord
    0.23
    Activations Density 0.005%

    No Known Activations