© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 39666
    Prev
    Next
    INDEX
    Explanations

    numbers and locations

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron detects NBA basketball–specific metadata, i.e. team and player names, game events, dates, and statistical or numerical tokens.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     its
    -1.61
     it
    -1.42
     fútbol
    -1.37
     there
    -1.30
     इसकी
    -1.22
    使其
    -1.20
    Its
    -1.17
    頜
    -1.16
     đảm
    -1.13
     metálico
    -1.12
    POSITIVE LOGITS
     может
    1.64
    ğum
    1.42
     from
    1.38
    那么
    1.35
     może
    1.30
     like
    1.30
     mungkin
    1.30
    рант
    1.29
    jelaskan
    1.28
    breviations
    1.27
    Activations Density 0.120%

    No Known Activations