© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 26449
    Prev
    Next
    INDEX
    Explanations

    digits and colons

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

    The main thing this neuron does is spot numeric digit tokens or contiguous digit sequences (especially zeros) in code or 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
     pidió
    -1.05
    쩌
    -1.02
    selection
    -0.94
    zow
    -0.94
     tentunya
    -0.90
     revient
    -0.90
    ≈
    -0.89
     selection
    -0.89
     \
    -0.88
     solicitó
    -0.86
    POSITIVE LOGITS
     المدار
    1.04
    遑
    1.01
    jestel
    1.00
    ěry
    0.98
    OCCURRED
    0.96
    ằm
    0.95
     time
    0.94
    ズニー
    0.92
     girly
    0.92
    minecraft
    0.91
    Activations Density 0.004%

    No Known Activations