© 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. 107170
    Prev
    Next
    INDEX
    Explanations

    document segments

    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
    æĹ¢æĺ¯
    -0.60
    åIJĮæĹ¶ä¹Łæĺ¯
    -0.51
    åIJĮæĹ¶ä¹Ł
    -0.49
    æĹ¢
    -0.46
    ä¸įä»ħæĺ¯
    -0.45
    æĹ¢èĥ½
    -0.44
    ä¸įä»ħèĥ½
    -0.43
    åħ¼åħ·
    -0.42
    åIJĮæŶ
    -0.41
     simultaneously
    -0.40
    POSITIVE LOGITS
    ensation
    0.29
     AU
    0.26
    工夫
    0.26
     outside
    0.26
     Yates
    0.25
     Surely
    0.25
    æĢİä¹ĪåĬŀ
    0.24
    èĪį
    0.24
     wo
    0.24
    å¹´åĪĿ
    0.24
    Activations Density 61.700%

    No Known Activations