© 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. 63970
    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
     exc
    -0.26
    çķĮ
    -0.26
    æľĢå°ij
    -0.25
    è¿ĩ
    -0.24
     retro
    -0.24
    æ¯Ĵ
    -0.24
    å¦ĥ
    -0.24
    etro
    -0.24
    çŁ¿
    -0.24
    è¿ĩå¾Ĺ
    -0.24
    POSITIVE LOGITS
    esign
    0.37
    iesen
    0.29
    æ°´éĿ¢
    0.27
    åĵģçīĮ形象
    0.26
    è¦ģåıĬæŶ
    0.26
    é¼ĭ
    0.25
     DbType
    0.25
    Routine
    0.25
    pter
    0.25
    å¼ķ导
    0.25
    Activations Density 0.049%

    No Known Activations

    This feature has no known activations.