© 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. 34-GEMMASCOPE-RES-131K
    7. 81409
    Prev
    Next
    INDEX
    Explanations

    naming places and features

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

    This neuron fires on the tokens in formal “name-has-been-proposed-by” naming statements—i.e. it picks up the phrase pattern around “the name which has been proposed by.”

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    んだよね
    -0.74
     Umwel
    -0.73
     bodyguard
    -0.71
    zsche
    -0.71
     Gefühl
    -0.70
    binant
    -0.69
     виде
    -0.68
     bể
    -0.68
    Рисунок
    -0.68
    分に
    -0.67
    POSITIVE LOGITS
     naming
    1.24
     names
    1.17
    Feature
    1.08
     feature
    1.06
    names
    1.05
     Gazetteer
    1.05
     FEATURE
    1.03
    naming
    1.02
    Names
    1.01
     name
    1.00
    Activations Density 0.039%

    No Known Activations