© 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. 30-TRANSCODER-HP
    7. 95858
    Prev
    Next
    INDEX
    Explanations

    reading

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    elsea
    -0.26
    åĮ»åѦéĻ¢
    -0.25
    Fixture
    -0.25
    è¿Ļä¸ĢåĪ»
    -0.24
    éĤ£ä¸ĢåĪ»
    -0.24
     cru
    -0.24
     programmed
    -0.24
    ä¿®
    -0.24
    lige
    -0.24
     Programm
    -0.23
    POSITIVE LOGITS
     reading
    0.50
    读
    0.48
    reading
    0.46
    read
    0.45
     read
    0.44
    READ
    0.43
    Reading
    0.43
     theory
    0.42
     Reading
    0.41
     THEORY
    0.41
    Activations Density 0.885%

    No Known Activations