© 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. 34400
    Prev
    Next
    INDEX
    Explanations

    page numbers, years, and citations

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

    The neuron strongly activates on bibliographic reference metadata—numbers (years, page/volume numbers), publisher names, and similar citation tokens.

    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
     mouseY
    -1.13
    mainAxisSize
    -1.08
    CHELL
    -1.06
    EdgeInsets
    -1.06
    秶
    -1.05
    Conclusão
    -1.05
    Website
    -1.02
     СИ
    -0.99
     januari
    -0.98
    Fprintf
    -0.98
    POSITIVE LOGITS
     depósitos
    1.04
    hinh
    1.00
     wohnung
    0.96
    ppure
    0.95
     olsa
    0.93
     vuelos
    0.92
     machte
    0.91
    흘
    0.88
     focused
    0.87
     escaleras
    0.87
    Activations Density 0.017%

    No Known Activations