© 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. 129815
    Prev
    Next
    INDEX
    Explanations

    This neuron fires on the main subject nouns (the entities like “government,” “students,” “customers,” “agents,” “I,” etc.) that head a clause.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    subjects and states

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     infatti
    -1.16
     in
    -1.03
    餃子
    -1.02
    は
    -1.02
    there
    -0.98
     tersebut
    -0.96
     there
    -0.93
    isiä
    -0.88
    的是
    -0.88
     those
    -0.87
    POSITIVE LOGITS
     sentimiento
    1.11
     nikt
    1.09
     will
    1.05
    pecabe
    1.03
    steamcommunity
    1.02
     Cette
    1.02
    chieben
    1.02
    
    0.98
    為に
    0.98
     μόνο
    0.98
    Activations Density 0.039%

    No Known Activations