© 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. 35-TRANSCODER-HP
    7. 40640
    Prev
    Next
    INDEX
    Explanations

    gibberish or names

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_35.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.35.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    å©¿
    -0.30
    unu
    -0.28
    ÑĨÑĥ
    -0.27
    swick
    -0.26
    ç¬ı
    -0.26
    éĽĨå¸Ĥ
    -0.26
     ilma
    -0.26
    ึà¸ģ
    -0.25
    -trash
    -0.25
    uste
    -0.25
    POSITIVE LOGITS
     Dop
    0.88
     bor
    0.79
     dop
    0.77
     Bor
    0.77
     pom
    0.75
    bor
    0.74
     Pom
    0.71
    Bon
    0.69
     Pon
    0.69
     Bon
    0.68
    Activations Density 81.191%

    No Known Activations