© 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. 128517
    Prev
    Next
    INDEX
    Explanations

    The main thing this neuron does is detect the “=” symbol in mathematical or code-like expressions.

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

    sequence and contrast markers

    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
     ideen
    -1.27
    TextEditing
    -1.09
    narod
    -1.07
     داشتند
    -1.06
     others
    -1.06
     kvin
    -1.03
     your
    -1.02
     engelsk
    -0.99
    cedures
    -0.99
     ".";
    -0.99
    POSITIVE LOGITS
     then
    1.61
     vervolgens
    1.20
    Then
    1.19
    Entonces
    1.13
     Kese
    1.10
     However
    1.07
     entanto
    1.05
     теперь
    1.02
    Luego
    1.00
     Then
    1.00
    Activations Density 0.001%

    No Known Activations