© 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. 25-TRANSCODER-HP
    7. 107071
    Prev
    Next
    INDEX
    Explanations

    root

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    士
    -0.28
    ipsis
    -0.27
    neys
    -0.25
    flen
    -0.25
    å¾Ģåīį
    -0.24
    tas
    -0.24
    èľĢ
    -0.24
    _drv
    -0.24
    icides
    -0.24
     Zam
    -0.24
    POSITIVE LOGITS
     square
    0.34
    reme
    0.32
     cause
    0.32
    square
    0.28
    èĮİ
    0.27
    AdapterFactory
    0.27
    Square
    0.26
     tongue
    0.26
    æħĪ
    0.26
     SQUARE
    0.26
    Activations Density 1.894%

    No Known Activations