© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 93142
    Prev
    Next
    INDEX
    Explanations

    code syntax markers

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

    The neuron fires on occurrences of the “while” loop construct (i.e. the keyword “while” and its loop condition).

    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
     already
    -1.14
     from
    -1.06
     już
    -0.98
    že
    -0.94
     tillegg
    -0.94
     allerede
    -0.94
     which
    -0.92
    畏
    -0.90
     affect
    -0.90
     who
    -0.89
    POSITIVE LOGITS
     weiteren
    0.90
    INFORMATION
    0.89
     bmi
    0.88
    ]="
    0.80
    Ch
    0.79
     Discogs
    0.79
     дальней
    0.77
     chanteuse
    0.76
     القص
    0.76
    étisme
    0.75
    Activations Density 0.001%

    No Known Activations