© 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. 39549
    Prev
    Next
    INDEX
    Explanations

    digits followed by hyphens

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

    The neuron is selectively detecting standalone numeric tokens (digits or number fragments) within the text.

    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
    jenih
    -1.15
    ことも
    -1.09
    えっ
    -1.00
     também
    -0.98
     superfí
    -0.98
    ragamo
    -0.94
    おわりに
    -0.94
    也没
    -0.93
    すっ
    -0.93
     โดย
    -0.92
    POSITIVE LOGITS
     as
    1.18
     at
    1.02
     another
    1.02
     and
    1.00
    tuvo
    0.98
     part
    0.91
     with
    0.90
    了啊
    0.90
     an
    0.89
    راحی
    0.89
    Activations Density 0.010%

    No Known Activations