© 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. 22-TRANSCODER-HP
    7. 111927
    Prev
    Next
    INDEX
    Explanations

    Incomplete sentences; "then"

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    crease
    -0.27
     unsuccessful
    -0.25
    vidence
    -0.25
    æĸ¡
    -0.25
    åľ¨èģĮ
    -0.25
    ldb
    -0.25
    creasing
    -0.25
    eltas
    -0.24
    ponsor
    -0.24
     myself
    -0.24
    POSITIVE LOGITS
    çĦ¶åIJİ
    0.26
     THEN
    0.25
     Suddenly
    0.25
    ?action
    0.25
    Suddenly
    0.24
    éļıåIJİ
    0.24
    THEN
    0.24
    OrDefault
    0.24
    以åIJİ
    0.24
    ä¸ĭä¸ĢæŃ¥
    0.24
    Activations Density 0.007%

    No Known Activations