© 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. 89309
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    Wnd
    -0.26
    ments
    -0.25
     numbering
    -0.25
    åIJĪéĩij
    -0.23
    vement
    -0.23
    trait
    -0.23
    tie
    -0.23
    éº»æľ¨
    -0.23
     Clayton
    -0.23
    ABCDEFG
    -0.23
    POSITIVE LOGITS
    oola
    0.28
    ported
    0.26
     Tradable
    0.26
     pé
    0.26
    ourg
    0.26
     pó
    0.26
     perg
    0.25
    ieran
    0.25
    inka
    0.24
    éļ¾æ°ij
    0.24
    Activations Density 0.115%

    No Known Activations

    This feature has no known activations.