© 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. 27747
    Prev
    Next
    INDEX
    Explanations

    reference points for comparison

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

    The main thing this neuron does is detect numeric expressions (e.g. numbers, percentages, and statistical values) 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_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
     aproximativ
    -1.04
     δύο
    -0.99
     molte
    -0.96
    Ꮹ
    -0.95
     deux
    -0.94
     mnoho
    -0.93
    must
    -0.92
    聼
    -0.92
     molti
    -0.92
     astonishing
    -0.91
    POSITIVE LOGITS
     at
    1.14
     baseline
    1.04
     reference
    1.03
    jadinya
    0.99
     which
    0.98
     using
    0.98
     before
    0.96
    parate
    0.94
     June
    0.90
     each
    0.90
    Activations Density 0.032%

    No Known Activations