© 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. 17733
    Prev
    Next
    INDEX
    Explanations

    code blocks

    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
    pane
    -0.29
    éĿ¢æĿ¿
    -0.28
    fram
    -0.26
     scram
    -0.26
    çŃIJ
    -0.26
    天èµĭ
    -0.25
    çİ°è´§
    -0.24
     Panel
    -0.24
    æ³¢çī¹
    -0.24
    帧
    -0.24
    POSITIVE LOGITS
     block
    0.43
    block
    0.41
     code
    0.41
     BLOCK
    0.41
    _block
    0.40
    -block
    0.40
    Block
    0.40
    Code
    0.40
     compound
    0.39
    _code
    0.39
    Activations Density 26.358%

    No Known Activations