© 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. 19510
    Prev
    Next
    INDEX
    Explanations

    say Korean last names

    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
    ìł¤
    -0.31
    VisualStyle
    -0.29
    æµİ
    -0.27
    å¾·å°Ķ
    -0.26
    æķ°
    -0.26
     midd
    -0.26
    sı
    -0.25
    É©
    -0.25
    WG
    -0.24
    éļIJç§ģ
    -0.24
    POSITIVE LOGITS
     Park
    0.30
    (Uri
    0.29
    Kim
    0.29
    åħ¨æĸĩ
    0.27
    Park
    0.27
     Kim
    0.27
    icter
    0.25
     Hong
    0.25
    åŁ¤
    0.25
     kim
    0.25
    Activations Density 0.131%

    No Known Activations