© 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. 19-TRANSCODER-HP
    7. 8355
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_19.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.19.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    è¶Ĭæĺ¯
    -0.28
    sumer
    -0.26
    eous
    -0.26
    äºĮ级
    -0.26
    iect
    -0.26
    verte
    -0.25
    éĢļé£İ
    -0.25
    æĴĩ
    -0.25
    è¦ģåľ¨
    -0.25
    yme
    -0.25
    POSITIVE LOGITS
    transparent
    0.27
    ucked
    0.26
    ç»ĵ
    0.26
    Transparent
    0.25
     transparent
    0.25
     Transparent
    0.25
    æį®è¯´
    0.24
    åıĺå½¢
    0.24
     falsely
    0.24
    /false
    0.24
    Activations Density 1.123%

    No Known Activations

    This feature has no known activations.