© 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. 7243
    Prev
    Next
    INDEX
    Explanations

    programming constructs like class, protected, function, import

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

    The neuron fires on programming‐language assignment syntax (e.g. “=”, array brackets, annotation parameters) indicating variable or property value assignments.

    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
    "):
    -0.99
     Irlanda
    -0.96
    🛃
    -0.87
    tieren
    -0.85
     Polisi
    -0.85
     Fakta
    -0.85
     feveral
    -0.84
     alin
    -0.83
     lavadora
    -0.82
     nephew
    -0.82
    POSITIVE LOGITS
     protected
    1.02
     his
    0.98
    protected
    0.82
    della
    0.80
     fill
    0.80
    āb
    0.79
     protect
    0.77
     including
    0.77
    āci
    0.76
     our
    0.76
    Activations Density 0.004%

    No Known Activations