© 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. 14887
    Prev
    Next
    INDEX
    Explanations

    a before counts of two

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

    The neuron fires on colon punctuation that introduces lists, definitions, or labeled sections (e.g. “parts:”, “rounds:”, “:”).

    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
     reemplazar
    -1.13
     possible
    -1.07
     szczególnie
    -1.02
     całą
    -1.02
     necesitar
    -1.02
     funzionalità
    -1.02
     começar
    -1.01
    bbies
    -1.00
     niektó
    -0.99
     szczegól
    -0.99
    POSITIVE LOGITS
     a
    1.51
     one
    1.39
    一个
    1.30
    一个是
    1.19
     another
    1.14
     an
    1.12
    一個
    1.07
    Цитата
    1.07
     एक
    1.05
    还有一个
    1.04
    Activations Density 0.170%

    No Known Activations