© 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. 40289
    Prev
    Next
    INDEX
    Explanations

    mathematical sequences and code variables

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

    This neuron detects index‐style notation (e.g. array indices or subscripts such as “i” in s[i] or subscripts like pₙ).

    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
    ningss
    -0.94
     iniziale
    -0.90
    matig
    -0.90
    Dossier
    -0.88
    সূত্র
    -0.88
    neming
    -0.85
     dosa
    -0.85
    Né
    -0.83
    بال
    -0.82
    isesi
    -0.82
    POSITIVE LOGITS
     to
    1.02
    but
    0.88
    Ⓟ
    0.83
    ⬜
    0.83
     other
    0.80
    Anything
    0.79
    ciutto
    0.77
     anything
    0.75
    ltd
    0.75
    ℝ
    0.75
    Activations Density 0.010%

    No Known Activations