© 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. 10-GEMMASCOPE-RES-131K
    7. 103003
    Prev
    Next
    INDEX
    Explanations

    HTML structure and content

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

    HTML/XML tags and their closing angle brackets in code or markup.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron fires on standalone numeric tokens (particularly numbers and decimal values) in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     preventivo
    -1.41
     duque
    -1.31
    ཥ
    -1.30
     ་
    -1.29
    뀨
    -1.27
     последу
    -1.26
    Ook
    -1.25
    Mijn
    -1.23
     thre
    -1.23
    pherds
    -1.22
    POSITIVE LOGITS
     Since
    1.70
     While
    1.49
     when
    1.42
     Because
    1.37
     because
    1.30
     Although
    1.29
     memutus
    1.24
     они
    1.22
    Since
    1.20
    ichtete
    1.19
    Activations Density 0.011%

    No Known Activations