© 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. 64220
    Prev
    Next
    INDEX
    Explanations

    Contains numbers

    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
    uitive
    -0.28
    orable
    -0.27
    ampa
    -0.27
    urname
    -0.26
    è°·
    -0.26
    ovid
    -0.25
    odium
    -0.25
    è¨Ģ
    -0.25
     patched
    -0.25
    rox
    -0.24
    POSITIVE LOGITS
    å½ĵéĢī
    0.25
    æĮģæľīçļĦ
    0.25
    æĮģæľī
    0.25
    åĿIJ
    0.24
     Holder
    0.24
    éģĹçķĻ
    0.24
    å°¼æĸ¯
    0.23
    sse
    0.23
    isters
    0.23
     holding
    0.23
    Activations Density 5.752%

    No Known Activations