© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 34112
    Prev
    Next
    INDEX
    Explanations

    resource identifiers

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

    The neuron responds to numeric literals (digits) in code or markup, i.e. tokens representing numbers or numeric constants.

    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
     очередь
    -1.34
    chó
    -1.23
    чи
    -1.16
    プ
    -1.14
     ekster
    -1.13
    МЕ
    -1.13
     также
    -1.11
     brillant
    -1.10
    мо
    -1.09
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    -1.09
    POSITIVE LOGITS
     help
    1.45
     now
    1.45
     okay
    1.35
     one
    1.33
     actually
    1.32
     second
    1.22
     things
    1.21
     strana
    1.20
     different
    1.20
     not
    1.19
    Activations Density 0.024%

    No Known Activations