© 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. 68600
    Prev
    Next
    INDEX
    Explanations

    conditional logic and code events

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

    This neuron recognizes programming‐language syntax tokens (e.g. parentheses, braces, keywords) to detect code snippets.

    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
     verändert
    -1.05
    oos
    -1.00
     parfois
    -0.99
    même
    -0.96
     reager
    -0.94
    σα
    -0.93
    都需要
    -0.93
    觊
    -0.91
    ванные
    -0.90
    vaní
    -0.89
    POSITIVE LOGITS
     but
    1.20
    Angeb
    1.03
     leaving
    1.02
     pObj
    0.99
     giving
    0.91
     spiaggia
    0.91
    Vorte
    0.90
    ではでは
    0.90
     during
    0.90
     rozwią
    0.90
    Activations Density 0.002%

    No Known Activations