© 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. 34-GEMMASCOPE-RES-131K
    7. 117848
    Prev
    Next
    INDEX
    Explanations

    Chinese names and disambiguation

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

    The neuron is picking out non‐word tokens—primarily numeric strings (e.g. years, codes) and adjacent punctuation marks—rather than ordinary vocabulary.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     majú
    -0.79
    ylon
    -0.77
     interesa
    -0.75
    ྲ
    -0.75
    Liquidity
    -0.74
     obstante
    -0.71
     sensibilité
    -0.70
    Cerca
    -0.70
    dcterms
    -0.69
    Juno
    -0.69
    POSITIVE LOGITS
    век
    0.77
     peel
    0.77
    ApiException
    0.74
     Pente
    0.71
    篇
    0.71
     respon
    0.70
    transformation
    0.69
    ϕ
    0.69
     Hoboken
    0.68
     transformation
    0.67
    Activations Density 0.012%

    No Known Activations