© 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. Qwen3-1.7B
    5. 27-LLAMASCOPE-2-LORSA-16K-K64
    7. 14504
    Prev
    Next
    INDEX
    Explanations

    say "three"

    unknown · unknown
    New Auto-Interp
    Top Features by Cosine Similarity
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    八个
    -23.38
    七个
    -20.63
    六个
    -18.75
    五个
    -17.75
     eight
    -15.69
    四个
    -14.75
    六大
    -14.06
    8
    -13.88
     seven
    -13.81
    7
    -13.75
    POSITIVE LOGITS
    3
    36.25
    ３
    31.38
     thirty
    29.50
     Three
    29.00
    -three
    28.50
     Thirty
    28.25
    三十
    27.88
    三百
    27.63
    三十年
    27.38
     three
    27.13
    Activations Density 0.488%

    No Known Activations

    This feature has no known activations.