© 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. 491
    Prev
    Next
    INDEX
    Explanations

    admission, lease, allocation, sold, sale

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

    The neuron is tuned to detect numeric tokens and numeral-rich strings (e.g., numbers, years, counts, and codes) in 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_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
    ག
    -0.78
    asure
    -0.74
    strates
    -0.73
    eniem
    -0.72
    uten
    -0.70
    beur
    -0.69
    NgModule
    -0.68
    bereich
    -0.66
    ciales
    -0.66
    esis
    -0.66
    POSITIVE LOGITS
     counseling
    1.91
     counselling
    1.87
     Counseling
    1.72
     COUNSEL
    1.52
     Counselling
    1.48
    Coun
    1.30
    coun
    1.03
     counselor
    0.96
     counselors
    0.94
    COUN
    0.93
    Activations Density 0.011%

    No Known Activations