© 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. 108577
    Prev
    Next
    INDEX
    Explanations

    :

    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
    å´§
    -0.26
    iode
    -0.26
     sworn
    -0.25
    èĿ
    -0.25
    ilent
    -0.25
    å®ŀçī©
    -0.25
     rented
    -0.25
     Enumerator
    -0.24
    agos
    -0.24
    frage
    -0.24
    POSITIVE LOGITS
    LK
    0.27
     meaning
    0.26
    ạc
    0.26
    rad
    0.25
    -.
    0.25
    asic
    0.24
    isis
    0.24
    Py
    0.24
    Rad
    0.24
     Py
    0.23
    Activations Density 1.587%

    No Known Activations