© 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. 10-GEMMASCOPE-RES-131K
    7. 127482
    Prev
    Next
    INDEX
    Explanations

    1950s years

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

    The neuron detects four-digit year tokens (i.e. date references like “1953,” “1954,” etc.).

    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
    戆
    -2.09
    接下来的
    -2.09
    鯔
    -2.06
    俛
    -1.99
    前面的
    -1.95
    鼱
    -1.90
     nevy
    -1.89
    Ꮵ
    -1.88
    鼯
    -1.87
    前来
    -1.86
    POSITIVE LOGITS
    1
    2.80
    2
    2.36
    </h3>
    2.20
    ，
    2.16
    '
    2.16
    ↵
    2.11
    .
    2.05
    3
    1.98
    ↵↵
    1.96
    8
    1.91
    Activations Density 0.195%

    No Known Activations