© 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. 30-TRANSCODER-HP
    7. 97008
    Prev
    Next
    INDEX
    Explanations

    God

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     rotated
    -0.27
    ynchronized
    -0.27
     instancia
    -0.26
     bred
    -0.25
    nid
    -0.25
    èIJ¥ä¸ļé¢Ŀ
    -0.25
    ä½ĵéªĮ
    -0.24
    stances
    -0.24
     branding
    -0.24
     stagger
    -0.24
    POSITIVE LOGITS
    (Button
    0.29
    å¾ĭ
    0.28
     eternity
    0.28
     hiá»ĩu
    0.26
     redeem
    0.25
    ä»ĸ对
    0.25
    æľĶ
    0.25
     errorCallback
    0.24
    åĬ«
    0.24
     Straw
    0.24
    Activations Density 0.893%

    No Known Activations