© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 117526
    Prev
    Next
    INDEX
    Explanations

    code separators and symbols

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

    The neuron strongly activates on numeric literals (digits) 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
     make
    -1.35
     made
    -1.28
     them
    -1.27
     dość
    -1.24
     focused
    -1.24
     like
    -1.23
     makes
    -1.20
     difficult
    -1.19
     not
    -1.16
     might
    -1.16
    POSITIVE LOGITS
      
    3.34
       
    2.30
        
    1.84
     restes
    1.28
    1.23
    graphiques
    1.20
     lada
    1.20
     EXCELL
    1.16
    maßnahmen
    1.16
     heer
    1.16
    Activations Density 0.078%

    No Known Activations