© 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. 141209
    Prev
    Next
    INDEX
    Explanations

    the

    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
    aked
    -0.28
    å·²ç»ıåľ¨
    -0.26
    xFD
    -0.25
     Sentry
    -0.25
    å·²ç»ıæĺ¯
    -0.24
     explic
    -0.24
    å¨ľ
    -0.24
    å·²ç»ıæľīäºĨ
    -0.23
     conspir
    -0.23
    毫ä¸į
    -0.23
    POSITIVE LOGITS
    çħ§
    0.26
    巨人
    0.25
    ä»Ģ
    0.25
     asses
    0.25
    =>
    0.24
    ÑĪение
    0.24
    rase
    0.24
     F
    0.24
     noises
    0.23
    åĻªéŁ³
    0.23
    Activations Density 22.046%

    No Known Activations