© 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. 130047
    Prev
    Next
    INDEX
    Explanations

    examples and specifications

    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
    VS
    -0.27
    acad
    -0.26
    -san
    -0.25
    =start
    -0.25
    -spe
    -0.24
     echoed
    -0.24
    æĬĺ
    -0.24
    çĶļ
    -0.23
    _mirror
    -0.23
     Mirror
    -0.23
    POSITIVE LOGITS
    ç»Ļèĩªå·±
    0.26
    sockopt
    0.26
    è¯ĦåĪĨ
    0.25
    eneg
    0.25
    对èĩªå·±çļĦ
    0.24
    åĽłä¸ºå¥¹
    0.24
     uninitialized
    0.24
    èī²ç´ł
    0.24
    uess
    0.23
     /*!
    0.23
    Activations Density 6.161%

    No Known Activations