© 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. 129505
    Prev
    Next
    INDEX
    Explanations

    erratic texts

    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
     Challenges
    -0.26
     mist
    -0.25
    highlight
    -0.25
    vido
    -0.24
     Horror
    -0.24
    rium
    -0.24
     last
    -0.24
    çīŁ
    -0.24
     temptation
    -0.24
    pir
    -0.24
    POSITIVE LOGITS
    nze
    0.28
    enga
    0.26
    YTE
    0.26
     delt
    0.25
    ï¼ļ"
    0.24
     dors
    0.24
    ERVE
    0.24
    eced
    0.23
     Northern
    0.23
    adamente
    0.23
    Activations Density 15.652%

    No Known Activations