© 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. 95418
    Prev
    Next
    INDEX
    Explanations

    string matching and searching

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

    The neuron fires on technical terms referring to sequences or subsequences in string‐processing contexts (e.g. “sequence,” “subsequence,” “string”).

    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
    áreas
    -0.94
    Associazione
    -0.88
     puits
    -0.88
     clasificación
    -0.87
    Élet
    -0.87
    jeux
    -0.84
    tratamiento
    -0.84
     różne
    -0.84
    ец
    -0.83
     vientre
    -0.82
    POSITIVE LOGITS
     smaller
    1.05
    月亮
    0.94
     shorter
    0.93
     pieces
    0.93
    你不要
    0.93
    Ϋ
    0.92
    azan
    0.91
    יפור
    0.90
     search
    0.90
     consisting
    0.90
    Activations Density 0.046%

    No Known Activations