© 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. 20-TRANSCODER-HP
    7. 141017
    Prev
    Next
    INDEX
    Explanations

    hyphens zeros

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    æijĩ头
    -0.28
    ApiClient
    -0.25
    adies
    -0.25
    quake
    -0.25
    ucci
    -0.24
     каÑĤал
    -0.24
    .setVertical
    -0.24
    æ½
    -0.24
    æ¶ķ
    -0.24
    Edited
    -0.24
    POSITIVE LOGITS
    好è¿IJ
    0.29
     World
    0.28
    ä¸ĢåĪĩ
    0.27
    ÑģÑĤÑĢо
    0.26
     Everything
    0.25
    æĿĤ
    0.25
    åIJĥäºı
    0.25
     early
    0.25
    DOC
    0.25
    izr
    0.25
    Activations Density 2.118%

    No Known Activations