© 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. 63009
    Prev
    Next
    INDEX
    Explanations

    Code and technical language

    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
    该å¦Ĥä½ķ
    -0.28
    etta
    -0.25
    约å®ļ
    -0.24
     Atl
    -0.24
    äºĭåħĪ
    -0.24
    æľĢåħĪ
    -0.23
     Plains
    -0.23
    å·¥ç¨ĭæĬĢæľ¯
    -0.23
     oldest
    -0.23
     followed
    -0.23
    POSITIVE LOGITS
    讹
    0.26
    /operator
    0.26
     enn
    0.25
    æѧè§Ĩ
    0.24
    OOD
    0.24
    åı·çº¿
    0.24
    nton
    0.23
    .Cond
    0.23
    ood
    0.23
    iator
    0.23
    Activations Density 71.783%

    No Known Activations