© 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. 9771
    Prev
    Next
    INDEX
    Explanations

    keywords and declarations

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

    The neuron is picking out numeric literals (numbers) 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_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
     indispens
    -1.29
     unités
    -1.27
     geweldig
    -1.23
    킁
    -1.23
     dinoss
    -1.21
     geweldige
    -1.21
    
    -1.20
    蔸
    -1.20
     vermelho
    -1.20
     Calidad
    -1.20
    POSITIVE LOGITS
     before
    1.21
     if
    1.17
     which
    1.14
     when
    1.13
     even
    1.10
     while
    1.10
     this
    1.09
     knows
    1.09
     during
    1.09
     as
    1.06
    Activations Density 0.021%

    No Known Activations