© 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. 130305
    Prev
    Next
    INDEX
    Explanations

    hypertension

    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
    erts
    -0.29
    åij¢
    -0.26
     pro
    -0.25
     quo
    -0.25
     simples
    -0.25
    åıįæŃ£
    -0.25
    ak
    -0.25
    æ¦Ĥ
    -0.24
    yl
    -0.24
    ïĤ
    -0.24
    POSITIVE LOGITS
    并ä¸Ķ
    0.29
    uish
    0.26
    dae
    0.26
    æ·±åħ¥
    0.26
    _into
    0.24
     Trash
    0.24
    bucks
    0.24
    ä»ĵåĤ¨
    0.23
    Altern
    0.23
    .realpath
    0.23
    Activations Density 34.963%

    No Known Activations