© 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. 50401
    Prev
    Next
    INDEX
    Explanations

    numbers and counts

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

    The neuron activates on cardinal numbers used in hyphenated constructs (e.g. “three-toed,” “four-legged,” “five-lined”).

    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
     bombas
    -0.86
     gestos
    -0.85
     晒
    -0.85
    绔
    -0.83
    殳
    -0.83
    TypeOf
    -0.83
    thirds
    -0.83
    安卓
    -0.82
     empezaron
    -0.82
    χος
    -0.82
    POSITIVE LOGITS
     seater
    0.89
     hole
    0.82
     day
    0.81
     osobow
    0.81
     wheel
    0.80
     compartment
    0.76
    person
    0.76
    継
    0.76
     actividad
    0.74
     thwart
    0.74
    Activations Density 0.055%

    No Known Activations