© 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. 34-GEMMASCOPE-RES-131K
    7. 109207
    Prev
    Next
    INDEX
    Explanations

    numbers and ranges

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

    The neuron is keyed to quantitative measurements, i.e. numerals and related numeric tokens (e.g. session counts, percentages, rest‐interval numbers, weights).

    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
     reps
    -1.00
    Reps
    -0.85
    WillAppear
    -0.82
     Reps
    -0.81
    èze
    -0.79
     اث
    -0.76
     testigo
    -0.76
     aeg
    -0.75
     Dois
    -0.75
    ics
    -0.73
    POSITIVE LOGITS
    Accessory
    0.85
     accessory
    0.81
    makeConstraints
    0.79
    밴
    0.77
    洗面
    0.77
     Bands
    0.77
     двига
    0.77
    quem
    0.75
    珊瑚
    0.74
    accessory
    0.72
    Activations Density 0.018%

    No Known Activations