© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 129784
    Prev
    Next
    INDEX
    Explanations

    The neuron triggers on isolated uppercase/math‐mode identifiers (single letters like “A,” “K,” “p,” “TM,” “Ad,” “N,” “CW,” etc.) in the text.

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

    $X$- or $X_$, $X($

    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
     allows
    -1.82
     them
    -1.63
     even
    -1.62
     thinks
    -1.59
     makes
    -1.59
     what
    -1.59
     بسیار
    -1.58
     give
    -1.55
     provides
    -1.55
     believes
    -1.51
    POSITIVE LOGITS
     faciles
    1.70
    '
    1.70
     difficiles
    1.70
    </em>
    1.59
     technische
    1.58
    ’
    1.50
     probablement
    1.45
     boden
    1.41
     baden
    1.38
     bestellen
    1.36
    Activations Density 0.007%

    No Known Activations