© 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. 96921
    Prev
    Next
    INDEX
    Explanations

    programming code

    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
     control
    -0.46
    æݧåĪ¶
    -0.42
    	control
    -0.38
    control
    -0.38
     Control
    -0.36
    /control
    -0.35
     condition
    -0.35
     flow
    -0.35
     controls
    -0.33
    -control
    -0.32
    POSITIVE LOGITS
    amage
    0.25
    imed
    0.25
    adir
    0.25
    mares
    0.24
     Sachs
    0.24
    erea
    0.24
    rawl
    0.24
    è°·
    0.24
    onymous
    0.24
    thur
    0.24
    Activations Density 0.905%

    No Known Activations