© 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. 107667
    Prev
    Next
    INDEX
    Explanations

    pool

    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
    uesta
    -0.28
     TW
    -0.27
     es
    -0.27
     screened
    -0.26
    screen
    -0.26
    ä»ĸæĺ¯ä¸Ģ个
    -0.26
    æĻĶ
    -0.25
     veget
    -0.25
     Frozen
    -0.24
     vaccinated
    -0.24
    POSITIVE LOGITS
    çķİ
    0.26
    irs
    0.26
    ä¹ĭä¸Ģ
    0.26
    æİĢ
    0.25
    巨头
    0.25
     terse
    0.25
    åıijåĩº
    0.25
    æŶä¸į
    0.25
    ä¹ĭ称
    0.24
    åİŁåĽłä¹ĭä¸Ģ
    0.24
    Activations Density 54.620%

    No Known Activations