© 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. 10-GEMMASCOPE-RES-131K
    4. 111169
    Prev
    Next
    INDEX
    Explanations

    be possible, be safe, be ok

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

    This neuron strongly activates on the verb “be” (especially in modal/question constructions like “would … be …,” “has … been …,” “are … be …,” etc.).

    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
    kurtka
    -2.02
    蠵
    -1.90
    '
    -1.88
     certainement
    -1.86
    
    -1.83
     miniaturka
    -1.80
    bluza
    -1.80
     marquées
    -1.80
     étudier
    -1.75
     Россия
    -1.71
    POSITIVE LOGITS
    </i>
    1.82
     then
    1.74
    他に
    1.70
    when
    1.70
     It
    1.67
    そういった
    1.64
     during
    1.63
    Not
    1.57
    ........
    1.55
          
    1.55
    Activations Density 0.102%

    No Known Activations