© 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. 68785
    Prev
    Next
    INDEX
    Explanations

    identifiers and paths

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

    The neuron is picking out code identifier tokens—especially camel-case or all-caps names and similar programming symbols.

    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
     humeur
    -0.96
     tristes
    -0.91
     sérieux
    -0.89
     čet
    -0.88
     BytesLike
    -0.88
    kozások
    -0.87
    ----------
    -0.85
     tropicales
    -0.84
     salario
    -0.84
    しましたが
    -0.83
    POSITIVE LOGITS
    </code>
    1.00
     гораздо
    0.99
    0.94
    0.91
    。
    0.89
                    
    0.88
     Djokovic
    0.85
    __
    0.84
     ngan
    0.82
     client
    0.81
    Activations Density 1.466%

    No Known Activations