© 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. 58451
    Prev
    Next
    INDEX
    Explanations

    layer of

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

    The neuron responds strongly to standalone numbers (numeric tokens) 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
     just
    -0.96
     gramas
    -0.94
     usually
    -0.94
     completely
    -0.93
     like
    -0.92
     so
    -0.90
     fairly
    -0.88
    useNewUrlParser
    -0.87
     most
    -0.86
     but
    -0.85
    POSITIVE LOGITS
        
    1.01
    bindet
    1.00
     proporcionan
    0.97
    졍
    0.97
     ULT
    0.92
     Cryptocurrency
    0.91
     rémun
    0.90
     côtés
    0.90
    nachweis
    0.90
     INCRE
    0.89
    Activations Density 0.001%

    No Known Activations