© 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
    2. Gemma-2-27B
    3. 34-GEMMASCOPE-RES-131K
    4. 59852
    Prev
    Next
    INDEX
    Explanations

    numbers followed by dates or text

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

    This neuron activates strongly on standalone numeric tokens (digits and simple numeric strings, including fractions).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     most
    -1.08
     very
    -1.02
     some
    -1.00
     small
    -0.97
     egiten
    -0.96
    蠅
    -0.94
     little
    -0.94
     diverse
    -0.93
     treno
    -0.91
     merely
    -0.90
    POSITIVE LOGITS
    >−</
    0.88
    0
    0.88
     longitudinally
    0.84
    fjspx
    0.82
     akumulator
    0.81
    )^{
    0.79
    copically
    0.77
     Legislature
    0.77
     boulevard
    0.76
    فوائد
    0.74
    Activations Density 0.008%

    No Known Activations