© 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. 25541
    Prev
    Next
    INDEX
    Explanations

    release gold, idTech, Kenzō, Q:

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

    The neuron fires on programming identifiers—especially CamelCase or capitalized names (e.g. variable, type, or contract names) 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
     religieux
    -1.72
     chré
    -1.70
     प्रोड
    -1.41
    ært
    -1.36
    髂
    -1.34
     تھے
    -1.30
     chrétien
    -1.27
     elevador
    -1.27
    艏
    -1.24
    Karma
    -1.21
    POSITIVE LOGITS
     menjalani
    1.48
     cubierto
    1.48
    𓂂
    1.42
     токси
    1.41
     каждого
    1.39
     الحكومة
    1.39
     proteínas
    1.38
     regalías
    1.37
    AFFIRMED
    1.36
    ateral
    1.34
    Activations Density 0.003%

    No Known Activations