© 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. 70363
    Prev
    Next
    INDEX
    Explanations

    many different types

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

    The neuron triggers on the formal, boilerplate connective phrases typical of technical or patent documents—e.g. “prior to,” “suggest that,” “defined in,” “many types of … have been developed,” etc.

    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
     two
    -2.58
     three
    -2.05
     another
    -1.77
     any
    -1.77
     or
    -1.68
    兩
    -1.65
    了两
    -1.63
     four
    -1.54
     some
    -1.50
     five
    -1.49
    POSITIVE LOGITS
     olika
    2.11
     различных
    2.03
     különböző
    1.97
     różnych
    1.95
     مختلف
    1.92
     различ
    1.87
     diferentes
    1.80
     verschiedenen
    1.78
     unterschied
    1.68
     différents
    1.66
    Activations Density 0.160%

    No Known Activations