© 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. 25-TRANSCODER-HP
    7. 108220
    Prev
    Next
    INDEX
    Explanations

    entering or being "on" something

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     attention
    -0.35
     conclusion
    -0.32
    åĨ³è®®
    -0.32
    表éĿ¢
    -0.32
     conclusions
    -0.30
    Surface
    -0.30
     surface
    -0.30
     Surface
    -0.30
    attention
    -0.29
    surface
    -0.28
    POSITIVE LOGITS
    鸥
    0.26
    PEC
    0.25
    invalidate
    0.24
    çº
    0.24
    fbe
    0.24
     Breast
    0.24
    conc
    0.23
    ches
    0.23
    å¦Ĭ
    0.23
    æĬķæĶ¾
    0.23
    Activations Density 1.077%

    No Known Activations