© 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. 31-TRANSCODER-HP
    7. 17803
    Prev
    Next
    INDEX
    Explanations

    commas and 2

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    ä¸Ģåı£æ°Ķ
    -0.31
    åįģåĪĨéĩįè¦ģ
    -0.28
    æīĵçł´
    -0.27
    ä¸Ģç»ı
    -0.27
    empo
    -0.26
    åĪ°æĿ¥
    -0.26
    å¿«æĿ¥
    -0.26
    prene
    -0.25
    ä¸Ģ大
    -0.25
    statement
    -0.25
    POSITIVE LOGITS
     indeed
    0.53
     above
    0.40
     Indeed
    0.40
    Indeed
    0.39
     sometimes
    0.39
    suggest
    0.37
    above
    0.37
     suggest
    0.37
     suggests
    0.36
     suggestions
    0.36
    Activations Density 1.650%

    No Known Activations