© 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. 19586
    Prev
    Next
    INDEX
    Explanations

    Names of people

    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
    Not in Any Lists

    No Comments

    Negative Logits
     qual
    -0.35
     dile
    -0.30
    çĽijå¯Ł
    -0.29
    æĮ½
    -0.28
    itim
    -0.27
     kron
    -0.25
    éĩıåĮĸ
    -0.25
    qual
    -0.25
     Kron
    -0.25
    å¼Ģ
    -0.24
    POSITIVE LOGITS
     Bid
    0.30
    zure
    0.28
    .$.
    0.27
    è¿Ľåıĸ
    0.27
    åİ©
    0.25
     Higgins
    0.25
    fullscreen
    0.25
    éĥİ
    0.24
     susceptible
    0.24
     bid
    0.24
    Activations Density 0.888%

    No Known Activations