© 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. 22-GEMMASCOPE-RES-131K
    7. 66085
    Prev
    Next
    INDEX
    Explanations

    version numbers and names

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

    The neuron predominantly responds to numeric tokens and numeric punctuation (digits, dates, years, and numbers formatted with commas or decimals).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    ymbol
    -1.00
    
    -0.98
    −/−
    -0.98
     panneaux
    -0.97
    correction
    -0.97
     danke
    -0.96
     defekt
    -0.96
     bomberos
    -0.94
     câbles
    -0.94
     ervaring
    -0.94
    POSITIVE LOGITS
     it
    1.17
     by
    1.03
     then
    0.95
    illen
    0.90
    ducteur
    0.88
     Some
    0.85
     that
    0.82
    的主要
    0.82
    ądze
    0.81
     one
    0.80
    Activations Density 0.206%

    No Known Activations