© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 22-TRANSCODER-HP
    7. 97136
    Prev
    Next
    INDEX
    Explanations

    Formal, academic writing

    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
    Negative Logits
     domestically
    -0.27
    WISE
    -0.26
    .getDeclared
    -0.25
    оÑĢо
    -0.24
    èĦ²
    -0.24
    пад
    -0.24
    éĴ§
    -0.24
    bd
    -0.24
     cud
    -0.24
     há»Ĺ
    -0.24
    POSITIVE LOGITS
     dia
    0.27
    红楼
    0.26
    dia
    0.26
    ika
    0.25
    ç¬Ķ
    0.25
     láºŃp
    0.25
    éŃģ
    0.25
    ifice
    0.24
    è¿Ļä¸Ģ天
    0.24
    eper
    0.24
    Activations Density 0.103%

    No Known Activations