© 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. 10-GEMMASCOPE-RES-131K
    7. 65163
    Prev
    Next
    INDEX
    Explanations

    mandates specific requirements

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

    The neuron preferentially activates on longer, content‐bearing tokens (e.g. multi‐letter words or symbols) rather than on common short function words or punctuation.

    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
     ”
    -1.71
     ‘
    -1.70
     “
    -1.57
     (
    -1.55
     (“
    -1.51
     A
    -1.45
     {
    -1.45
     ’
    -1.43
     «
    -1.38
     expresa
    -1.38
    POSITIVE LOGITS
     deudas
    1.52
     opportunities
    1.48
     OGSÅ
    1.42
     Locomotive
    1.38
    と思いますが
    1.36
    ౖ
    1.35
     啦
    1.35
     acogedor
    1.34
     costumi
    1.34
     kön
    1.34
    Activations Density 0.053%

    No Known Activations