© 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. 28391
    Prev
    Next
    INDEX
    Explanations

    unique identifiers or shared properties

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

    The neuron fires on identifier names for ID fields (e.g. “id,” “parent.id,” “group id”) in code or data descriptions.

    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
     zarówno
    -0.93
     such
    -0.90
     située
    -0.89
     néhány
    -0.85
    缚
    -0.84
     flexibilidad
    -0.84
     accueillir
    -0.83
    derm
    -0.82
     öff
    -0.81
     eth
    -0.80
    POSITIVE LOGITS
     different
    1.25
     shared
    1.22
     unique
    1.10
     identific
    1.08
     identifica
    1.07
    shared
    1.00
    同一个
    1.00
     sharing
    0.95
     identifier
    0.94
     numbering
    0.93
    Activations Density 0.102%

    No Known Activations