© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 10562
    Prev
    Next
    INDEX
    Explanations

    competition entrants

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

    The neuron is tuned to detect numeric tokens—years, dates, rankings, seed numbers, scores, and other digit-based information.

    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
    )
    -1.95
    _
    -1.93
    }
    -1.91
     I
    -1.77
    :
    -1.75
    /
    -1.70
     now
    -1.64
    </strong>
    -1.56
    <
    -1.55
     cuando
    -1.53
    POSITIVE LOGITS
     DELLE
    1.73
    Soorten
    1.70
    颚
    1.70
     аппети
    1.64
     either
    1.59
    México
    1.59
     Comprometido
    1.54
    Perú
    1.52
     suche
    1.50
    莚
    1.48
    Activations Density 0.011%

    No Known Activations