© 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. 22-GEMMASCOPE-RES-131K
    7. 98581
    Prev
    Next
    INDEX
    Explanations

    English words in lists

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

    This neuron selectively activates on numeric tokens, section and figure references, and other formal metadata markers (e.g. Roman numerals, numbered survey results, headings, and citation-style tokens).

    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
    鸨
    -0.94
    벚
    -0.85
    atile
    -0.85
    퐼
    -0.85
    ering
    -0.83
     misschien
    -0.83
     analogies
    -0.82
     テレビ
    -0.82
    ETIES
    -0.81
    μένη
    -0.81
    POSITIVE LOGITS
     other
    1.41
     elsewhere
    1.16
     another
    1.15
     this
    1.09
    另一
    1.08
     دیگری
    1.03
     similarly
    0.99
     otros
    0.98
     anderer
    0.98
     másik
    0.97
    Activations Density 0.158%

    No Known Activations