© 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. 33-TRANSCODER-HP
    7. 54704
    Prev
    Next
    INDEX
    Explanations

    science prefixes

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    entes
    -0.27
    év
    -0.26
     Nation
    -0.24
    èĬ¥
    -0.24
    Sentence
    -0.23
    è°¢
    -0.23
     "<?
    -0.23
    éĩĮçļĦ
    -0.23
    ogle
    -0.23
    _chunk
    -0.22
    POSITIVE LOGITS
    thern
    0.33
    ther
    0.31
    -On
    0.30
    -Th
    0.29
    -Ass
    0.29
    -Pro
    0.29
    _Arg
    0.28
     Therm
    0.27
    -In
    0.27
     Ser
    0.27
    Activations Density 0.030%

    No Known Activations