© 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. 10-GEMMASCOPE-RES-131K
    7. 60985
    Prev
    Next
    INDEX
    Explanations

    zero modifies or quantifies

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

    The neuron fires on occurrences of the word “zero,” including when it appears in hyphenated or compound forms.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     somewhat
    -2.27
    somewhat
    -2.23
    ↵↵
    -2.19
    涐
    -2.11
     Somewhat
    -1.86
     dismal
    -1.84
    少し
    -1.83
     decidedly
    -1.80
    slightly
    -1.74
    訫
    -1.74
    POSITIVE LOGITS
     descanso
    1.77
     decreased
    1.75
     any
    1.70
    كتشاف
    1.63
    !)
    1.63
    //
    1.62
     has
    1.62
    前面的
    1.61
     believes
    1.60
     wants
    1.59
    Activations Density 0.010%

    No Known Activations