© 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. 95240
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    Not in Any Lists

    No Comments

    Negative Logits
     ngại
    -0.30
     time
    -0.29
    åħij
    -0.29
    time
    -0.28
    æŶéĹ´
    -0.28
    æŶéĹ´åĨħ
    -0.28
    åĵĤ
    -0.27
     Romanian
    -0.27
    $time
    -0.26
    -time
    -0.26
    POSITIVE LOGITS
    pg
    0.26
    ellow
    0.25
    /cpu
    0.24
    bez
    0.24
     grips
    0.24
    phon
    0.24
    (pg
    0.23
     Discipline
    0.23
    åħ³
    0.23
    ->{$
    0.23
    Activations Density 0.935%

    No Known Activations

    This feature has no known activations.