© 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. 22-GEMMASCOPE-RES-131K
    7. 13648
    Prev
    Next
    INDEX
    Explanations

    reaching a limit or value

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

    The neuron fires on numerical values (often with words like “reaches,” “hits,” or units) that mark specific size or quantity thresholds.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     COMMUNIC
    -1.08
     ersche
    -1.02
     jati
    -1.01
     maniere
    -1.00
    frauen
    -0.99
     poud
    -0.99
     ECONOMIC
    -0.98
     poë
    -0.98
    饃
    -0.96
     marchandises
    -0.95
    POSITIVE LOGITS
     become
    1.53
     becomes
    1.45
     or
    1.30
     began
    1.17
     begin
    1.17
     дости
    1.10
     begins
    1.06
     trở
    1.05
     became
    1.04
     reached
    1.02
    Activations Density 0.046%

    No Known Activations