© 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. 40189
    Prev
    Next
    INDEX
    Explanations

    numbers and specific tokens

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

    The neuron activates on long sequences of digits (often parts of catalog identifiers or coordinate-like numeric strings).

    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
     pantal
    -1.28
     vols
    -1.27
     theolog
    -1.13
     Aé
    -1.09
     sapat
    -1.08
     boute
    -1.06
     Giochi
    -1.05
    ほら
    -1.04
     regata
    -1.03
     vinho
    -1.02
    POSITIVE LOGITS
    <bos>
    1.30
    dü
    1.29
    pergillus
    1.19
     barvy
    1.17
    ลอง
    1.15
     búsquedas
    1.13
    Proč
    1.12
    始める
    1.11
     rengi
    1.11
     AMAZ
    1.08
    Activations Density 0.014%

    No Known Activations