© 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. 22-TRANSCODER-HP
    7. 86748
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     Morr
    -0.25
    Asia
    -0.25
    -ra
    -0.25
    /XML
    -0.24
    )NULL
    -0.24
    INET
    -0.23
    -China
    -0.23
     hangs
    -0.23
    arrants
    -0.23
     Quarry
    -0.23
    POSITIVE LOGITS
    åīįçŀ»æĢ§
    0.28
    bage
    0.27
    æ§Ľ
    0.25
    ç²¹
    0.25
     pretext
    0.24
    åŃ¦æľ¯
    0.24
    ack
    0.24
    é»ijçĻ½
    0.24
    ÅĽni
    0.24
    _util
    0.23
    Activations Density 0.021%

    No Known Activations

    This feature has no known activations.