© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 33-TRANSCODER-HP
    7. 52617
    Prev
    Next
    INDEX
    Explanations

    recognize/recognise/recognizing

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     recognition
    -0.74
     Recognition
    -0.67
     recognise
    -0.67
     recognize
    -0.66
     recognised
    -0.65
     recogn
    -0.65
    recogn
    -0.62
     recognizing
    -0.61
     recognizes
    -0.61
     recognized
    -0.60
    POSITIVE LOGITS
     and
    0.40
    åĴĮ
    0.34
    æĪı
    0.32
     the
    0.31
     both
    0.28
     nas
    0.27
    åĽ½æľīèµĦ产
    0.26
    缮
    0.25
     åĴĮ
    0.24
    åIJĮæł·çļĦ
    0.24
    Activations Density 0.027%

    No Known Activations