© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 18-TRANSCODER-HP
    7. 89468
    Prev
    Next
    INDEX
    Explanations

    Text with mixed symbols

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    åħ¨æĸĩ
    -0.30
    bage
    -0.27
    istine
    -0.26
    åºŀ
    -0.26
     reached
    -0.25
    æ°¸
    -0.25
    çľī
    -0.25
    realm
    -0.25
    å°½
    -0.25
    .localization
    -0.24
    POSITIVE LOGITS
    ç»ĦåIJĪ
    0.30
    è²ī
    0.29
    .compose
    0.28
    combine
    0.27
     combine
    0.27
    åIJĪæĪIJ
    0.27
     Combine
    0.26
    _concat
    0.26
    ga
    0.26
    compose
    0.26
    Activations Density 11.030%

    No Known Activations