© 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. 27-TRANSCODER-HP
    7. 9081
    Prev
    Next
    INDEX
    Explanations

    Frank

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_27.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.27.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    icator
    -0.25
    æĪĮ
    -0.25
    vak
    -0.24
    ç«ĭä½ĵ
    -0.24
     undert
    -0.24
    icut
    -0.24
     mundial
    -0.24
    sar
    -0.24
    ä¸ŃåŃĺåľ¨çļĦ
    -0.23
     Whereas
    -0.23
    POSITIVE LOGITS
    oen
    0.30
    sgiving
    0.29
    lier
    0.26
    好çľĭ
    0.26
    è¯Ŀ
    0.25
     Noon
    0.25
    ovsky
    0.25
    izations
    0.24
    formats
    0.24
    otron
    0.24
    Activations Density 1.052%

    No Known Activations