© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 124733
    Prev
    Next
    INDEX
    Explanations

    expressed by, confirmed, relates to, heights earlier

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

    The neuron fires on longer, multi-syllable content words—especially longer verbs and nouns—rather than on short function words.

    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
     навіть
    -1.78
     nawet
    -1.52
     даже
    -1.43
     тільки
    -1.39
     only
    -1.39
     even
    -1.38
     حتی
    -1.26
    ƌ
    -1.25
     ocurrió
    -1.24
     this
    -1.23
    POSITIVE LOGITS
     a
    1.83
    袤
    1.45
     way
    1.36
     kind
    1.34
     present
    1.30
     feel
    1.28
     anderen
    1.28
     andere
    1.27
     shown
    1.24
     obr
    1.23
    Activations Density 0.159%

    No Known Activations