© 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. 39633
    Prev
    Next
    INDEX
    Explanations

    groups of people

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

    The neuron strongly activates on plural nouns naming groups or categories of people.

    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
     himself
    -1.09
    larda
    -1.05
     itself
    -0.97
     сам
    -0.96
     has
    -0.94
     forskj
    -0.90
     is
    -0.90
    獨
    -0.88
    独
    -0.88
     lille
    -0.85
    POSITIVE LOGITS
     themselves
    2.11
    themselves
    1.50
     their
    1.47
     they
    1.46
     brothers
    1.40
     collectively
    1.38
     duo
    1.32
     two
    1.31
     were
    1.25
     trio
    1.20
    Activations Density 0.180%

    No Known Activations