© 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. 20-TRANSCODER-HP
    7. 108709
    Prev
    Next
    INDEX
    Explanations

    digits

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    ewood
    -0.30
     capit
    -0.29
    anker
    -0.28
    rol
    -0.27
     mon
    -0.25
     pat
    -0.25
    pike
    -0.25
     Falk
    -0.25
     iam
    -0.25
    æĻĥ
    -0.24
    POSITIVE LOGITS
    ling
    0.29
    ldb
    0.29
    ~=
    0.26
    åıĪè¦ģ
    0.25
    è¦ģèĢĥèĻij
    0.25
    gien
    0.25
    usercontent
    0.24
    塵
    0.24
    rez
    0.24
    _guess
    0.24
    Activations Density 3.028%

    No Known Activations