© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25834
    Prev
    Next
    INDEX
    Explanations

    places and companies

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

    The neuron strongly activates on capitalized proper names—especially countries, sports teams, and major organizations—so it’s detecting mentions of named entities.

    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
     obé
    -1.30
     souvent
    -1.23
    eut
    -1.21
     décide
    -1.14
     pensez
    -1.12
    gunas
    -1.12
    儡
    -1.12
    閼
    -1.12
     encantador
    -1.12
    Ệ
    -1.10
    POSITIVE LOGITS
     can
    1.58
     could
    1.55
     will
    1.40
     only
    1.23
    tieth
    1.20
    teenth
    1.16
    man
    1.14
     $
    1.12
    omato
    1.10
     should
    1.07
    Activations Density 0.015%

    No Known Activations