© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 65223
    Prev
    Next
    INDEX
    Explanations

    concepts and relationships

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

    The neuron activates on words used in questions about comprehension and clarity—terms like understand, confusing, sense, follow, going on, and similar that probe whether something makes sense.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     generell
    -1.38
    šić
    -1.29
     deux
    -1.27
    oudoune
    -1.23
    Overall
    -1.23
    zicht
    -1.20
    ОН
    -1.17
     livres
    -1.16
     Februar
    -1.15
    Companies
    -1.15
    POSITIVE LOGITS
     about
    2.08
     that
    1.80
     the
    1.65
     what
    1.63
     things
    1.55
     those
    1.48
     people
    1.46
     quickly
    1.43
     asked
    1.43
     there
    1.41
    Activations Density 0.401%

    No Known Activations