© 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. 84376
    Prev
    Next
    INDEX
    Explanations

    infrequently or rarely

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

    The neuron fires on numeric tokens and measurements (e.g. counts, statistics, percentages, years), i.e. it detects numbers in the text.

    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
     many
    -0.96
     banyak
    -0.81
     different
    -0.79
    တ်
    -0.75
    ætte
    -0.75
    련
    -0.72
     diferente
    -0.70
     diferentes
    -0.70
     zunächst
    -0.69
    maksud
    -0.69
    POSITIVE LOGITS
     rarely
    6.72
    rare
    5.78
     rare
    5.53
    Rare
    5.31
     Rarely
    5.19
     infrequently
    4.94
     seldom
    4.88
     selten
    4.72
     RARE
    4.66
     редко
    4.63
    Activations Density 0.297%

    No Known Activations