© 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. 48392
    Prev
    Next
    INDEX
    Explanations

    actions and states

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

    The neuron flags content‐heavy, multi‐syllable verbs typical of formal or technical prose.

    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
     have
    -1.76
    only
    -1.66
     another
    -1.66
     tenho
    -1.60
     were
    -1.59
    ch
    -1.58
     only
    -1.57
     tecnologías
    -1.54
     effectively
    -1.52
     approximately
    -1.51
    POSITIVE LOGITS
    梱
    1.46
    sebagai
    1.45
    𝐇
    1.37
    gebouw
    1.34
    ₦
    1.33
    Михаил
    1.31
    他们
    1.31
     organis
    1.31
    Татьяна
    1.30
    Максим
    1.30
    Activations Density 0.159%

    No Known Activations