© 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. 130046
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    å¼Ĥ
    -0.30
    å±Ĥ
    -0.29
     shaken
    -0.29
     hos
    -0.28
    getic
    -0.28
    好象
    -0.28
    èįª
    -0.28
    èµĽåĮº
    -0.28
    è´¥
    -0.27
    APSHOT
    -0.27
    POSITIVE LOGITS
    ote
    0.30
     perfection
    0.29
    .locals
    0.29
     Universal
    0.28
    Universal
    0.27
    impact
    0.26
    TEL
    0.26
     values
    0.25
    ol
    0.25
     Writing
    0.25
    Activations Density 1.108%

    No Known Activations

    This feature has no known activations.