© 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. 22-GEMMASCOPE-RES-131K
    7. 98166
    Prev
    Next
    INDEX
    Explanations

    code block start and punctuation

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

    The neuron activates on numeric literal tokens (numbers) in code.

    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
     only
    -1.60
     their
    -1.19
     after
    -1.10
     this
    -1.09
     just
    -1.09
     first
    -1.09
     before
    -1.07
     unrelenting
    -1.01
     all
    -1.01
     both
    -1.00
    POSITIVE LOGITS
     vélos
    1.16
     recientemente
    1.06
     někter
    1.06
     desperdic
    1.05
     değiştir
    1.00
     jouets
    1.00
     élar
    1.00
    alquiler
    0.99
     lutar
    0.99
     rassemb
    0.98
    Activations Density 0.000%

    No Known Activations