© 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. 34-GEMMASCOPE-RES-131K
    7. 48228
    Prev
    Next
    INDEX
    Explanations

    SafeMath arithmetic and test cases

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

    The neuron strongly activates on tokens that are numeric literals (i.e. sequences of digits) in code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     vaca
    -0.87
     pomme
    -0.86
     sorri
    -0.84
     Zweck
    -0.81
     puni
    -0.80
     katun
    -0.79
     yong
    -0.79
     Tela
    -0.78
     papas
    -0.78
     silang
    -0.78
    POSITIVE LOGITS
    ChildScrollView
    0.99
    SafeMath
    0.93
     overflow
    0.92
     ari
    0.91
     
    0.90
     Formula
    0.89
    uint
    0.88
    гости
    0.84
    Formula
    0.84
     Uint
    0.84
    Activations Density 0.001%

    No Known Activations