© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 41713
    Prev
    Next
    INDEX
    Explanations

    calculated actions

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

    The main thing this neuron does is detect the various forms of “calculate” (e.g. calculate, calculated, calculation, calculating).

    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
     Thousands
    -2.19
     Ultimately
    -2.02
     Afterward
    -1.95
     Hundreds
    -1.89
     Technically
    -1.83
     Notably
    -1.81
     Nowadays
    -1.80
     Whether
    -1.78
     What
    -1.75
     Owing
    -1.75
    POSITIVE LOGITS
    臯
    1.74
     that
    1.74
     também
    1.66
    紲
    1.63
    th
    1.62
     Hinblick
    1.59
     aussi
    1.58
    </th>
    1.58
    1.55
    ),
    1.50
    Activations Density 0.019%

    No Known Activations