© 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. 118080
    Prev
    Next
    INDEX
    Explanations

    speech and actions

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

    The neuron activates strongly on numeric tokens (i.e., any standalone number 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
    摇摇头
    -1.15
     reconoció
    -1.07
    拍了拍
    -1.05
    低声
    -1.05
    摇了摇头
    -1.02
    轻声
    -1.01
    笑了笑
    -1.00
     zastos
    -0.99
     aparecen
    -0.99
    叹了口气
    -0.97
    POSITIVE LOGITS
     said
    3.00
     says
    2.22
     asked
    2.05
     spoke
    1.95
     ask
    1.85
     asks
    1.80
     speak
    1.73
     say
    1.67
     nói
    1.59
     saying
    1.56
    Activations Density 0.007%

    No Known Activations