© 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. 88744
    Prev
    Next
    INDEX
    Explanations

    Code and numbers

    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
    segue
    -0.26
    éĥ½ä¸įä¼ļ
    -0.25
    Concern
    -0.23
    гон
    -0.23
     frame
    -0.23
     Confeder
    -0.23
    bens
    -0.23
    æľªæĪIJ
    -0.23
    éĥ½æľīçĿĢ
    -0.23
    æľªç»ıæİĪæĿĥ
    -0.22
    POSITIVE LOGITS
    æĹ¢æĺ¯
    0.27
    rips
    0.27
     either
    0.26
    ä¸įå®ļ
    0.26
    ç«łç¨ĭ
    0.25
    æµ·éĩı
    0.25
    åıĬåħ¶ä»ĸ
    0.24
     Cobb
    0.24
    (ca
    0.23
    èĢģ大
    0.23
    Activations Density 7.736%

    No Known Activations