© 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. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 92850
    Prev
    Next
    INDEX
    Explanations

    code and configuration structure

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

    The neuron is firing most strongly on numeric tokens (literal numbers) in code, config, or documentation. It's essentially detecting numeric values wherever they appear.

    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
    DataContract
    -0.96
     maal
    -0.83
     poliester
    -0.82
    леный
    -0.79
     polie
    -0.77
     кровати
    -0.77
     farmaco
    -0.76
     comuni
    -0.75
     jari
    -0.75
     culturale
    -0.75
    POSITIVE LOGITS
     unset
    0.91
    налу
    0.82
     muck
    0.75
    其实是
    0.75
     Righ
    0.75
    ISTAN
    0.73
    ({
    0.71
    Լ
    0.71
    /**
    0.70
    ✟
    0.70
    Activations Density 0.002%

    No Known Activations