© 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. 13123
    Prev
    Next
    INDEX
    Explanations

    numbers and punctuation in technical contexts

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

    This neuron activates on long numeric identifier strings (e.g. multi‐digit sequences in DOIs, reference numbers, PACS codes, etc.).

    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
    部分を
    -0.96
    ULU
    -0.95
     heilige
    -0.94
     alcuni
    -0.91
     kazi
    -0.88
    mito
    -0.85
     alcune
    -0.85
     遇
    -0.84
     jednego
    -0.84
     aktuell
    -0.84
    POSITIVE LOGITS
    ән
    1.06
     at
    0.98
    Kval
    0.97
     précédente
    0.94
    wnętr
    0.94
    ))
    0.94
    )),
    0.94
    멕
    0.93
    diğim
    0.91
    Accordion
    0.91
    Activations Density 0.000%

    No Known Activations