© 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. 113072
    Prev
    Next
    INDEX
    Explanations

    all / any

    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
    ç³»
    -0.29
     great
    -0.29
     firm
    -0.27
     broken
    -0.26
    uz
    -0.26
    Vo
    -0.25
    elo
    -0.24
     fir
    -0.24
    uze
    -0.23
    าà¸ĩ
    -0.23
    POSITIVE LOGITS
    -slot
    0.29
    toa
    0.28
     Hubb
    0.26
    槽
    0.26
     Conveyor
    0.25
    daemon
    0.25
    tron
    0.25
     automáticamente
    0.25
    .Combine
    0.25
    -Men
    0.25
    Activations Density 0.023%

    No Known Activations