© 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. 130314
    Prev
    Next
    INDEX
    Explanations

    Code/technical letters and 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
     Wass
    -0.27
    cuts
    -0.27
    æ¤ģ
    -0.25
    ë°ķ
    -0.25
    :max
    -0.24
    iali
    -0.23
    è¶ħé¢Ŀ
    -0.23
    ä¸įçĦ¶
    -0.22
     saison
    -0.22
    nds
    -0.22
    POSITIVE LOGITS
    ogle
    0.27
     IGN
    0.26
    è°Ł
    0.26
     или
    0.26
     or
    0.25
    NEY
    0.25
    çļĦçľ¼ç¥ŀ
    0.24
    /ros
    0.24
     Chrome
    0.24
     convo
    0.24
    Activations Density 12.027%

    No Known Activations