© 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. 22-GEMMASCOPE-RES-131K
    7. 128453
    Prev
    Next
    INDEX
    Explanations

    The neuron is primarily firing on standalone numeric tokens (numbers) in the text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    sentence endings and questions

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     équipé
    -1.12
     professionali
    -1.09
    Soorten
    -1.06
     rác
    -1.02
     partire
    -1.02
    וּ
    -1.00
     cvt
    -1.00
     दोनों
    -0.99
     trabajado
    -0.98
     Ayah
    -0.98
    POSITIVE LOGITS
     all
    1.44
     etc
    1.34
     lastly
    1.33
     even
    1.30
     likewise
    1.27
     you
    1.23
     when
    1.22
     Ebenso
    1.16
     （
    1.16
     those
    1.13
    Activations Density 0.005%

    No Known Activations