© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 35197
    Prev
    Next
    INDEX
    Explanations

    programming and configuration

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

    The neuron fires on deeply indented lines in nested JSON or object literals, effectively spotting deeply nested key/value entries.

    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
    ).
    -1.07
     spô
    -0.88
    âteau
    -0.88
     Phú
    -0.87
    norme
    -0.85
    cocho
    -0.85
     spiegel
    -0.84
     špec
    -0.83
     instead
    -0.83
     różnych
    -0.83
    POSITIVE LOGITS
    écuter
    1.07
    ılacak
    1.05
    Synchronization
    1.02
     these
    1.00
    ástica
    0.98
     слад
    0.97
    .+\
    0.97
     umum
    0.95
     CYTHON
    0.93
     semblables
    0.93
    Activations Density 0.006%

    No Known Activations