© 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. 22-GEMMASCOPE-RES-131K
    7. 13438
    Prev
    Next
    INDEX
    Explanations

    prefixes or suffixes

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

    This neuron fires on ordinary words that immediately precede numeric values, i.e. tokens occurring right before 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_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
    Selamat
    -1.16
    がありますが
    -1.06
    文獻
    -1.05
     apparaissent
    -1.05
     accompagner
    -1.04
    Kepala
    -1.02
    lardır
    -1.01
    OBJETIVO
    -1.00
     quaisquer
    -1.00
    Cevap
    -1.00
    POSITIVE LOGITS
     would
    1.05
     \&
    0.98
    妮
    0.97
    }*/
    0.94
    
    0.94
    nicos
    0.93
     then
    0.92
    ₫
    0.91
     (\
    0.91
     or
    0.91
    Activations Density 0.057%

    No Known Activations