© 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. 66865
    Prev
    Next
    INDEX
    Explanations

    support for various technical terms

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

    The neuron is triggered by tokens that look like technical acronyms or identifiers—i.e. strings containing uppercase letters, digits, or punctuation (e.g. I2C, XPath, Unicode, *.pkl.gz).

    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
     about
    -1.20
    でしたか
    -1.09
     vezes
    -1.02
     posibles
    -0.99
    羊肉
    -0.96
    幸好
    -0.90
    ırs
    -0.88
    ையில்
    -0.88
    acter
    -0.88
    bores
    -0.88
    POSITIVE LOGITS
    Any
    1.19
    any
    1.11
     kapan
    1.06
     both
    1.05
    upto
    1.05
     através
    1.03
     türkiye
    1.03
    Anything
    1.02
     kaya
    1.02
    多种
    1.02
    Activations Density 0.054%

    No Known Activations