© 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. 19660
    Prev
    Next
    INDEX
    Explanations

    Physics papers and symmetry

    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
    æĭ±
    -0.35
    zug
    -0.31
    åŁº
    -0.29
    çIJĨ论
    -0.28
    rys
    -0.28
     theoretical
    -0.27
    bose
    -0.27
    ç¬Ķ
    -0.27
    StringUtils
    -0.26
    KeyType
    -0.26
    POSITIVE LOGITS
    asio
    0.28
    ohen
    0.28
     CActive
    0.28
    empor
    0.25
    æݧåĪ¶åĻ¨
    0.25
    antasy
    0.25
    raft
    0.24
     cons
    0.24
    ç³Ł
    0.24
    çłŃ
    0.24
    Activations Density 1.775%

    No Known Activations