© 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. 33-TRANSCODER-HP
    7. 54594
    Prev
    Next
    INDEX
    Explanations

    role

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
    ger
    -0.28
    il
    -0.28
    alli
    -0.27
    alytics
    -0.27
    andy
    -0.26
    otto
    -0.26
    ”?
    -0.26
     il
    -0.26
    inka
    -0.25
    cmc
    -0.25
    POSITIVE LOGITS
     roles
    0.62
     Roles
    0.57
    çļĦè§Ĵèī²
    0.56
     role
    0.55
    è§Ĵèī²
    0.52
    roles
    0.49
    -role
    0.48
     Role
    0.48
    role
    0.48
    _role
    0.47
    Activations Density 0.029%

    No Known Activations