© 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. 22-GEMMASCOPE-RES-131K
    7. 95792
    Prev
    Next
    INDEX
    Explanations

    existential or universal quantifiers

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

    This neuron detects parenthesized part labels (e.g. the “(a)”, “(b)”, etc.) that mark sub-questions in a math question.

    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
     intptr
    -1.23
     arquitecto
    -1.16
     капю
    -1.14
    haviours
    -1.12
     nossa
    -1.11
     embaj
    -1.10
    სქოლიო
    -1.09
    providedIn
    -1.09
     kelime
    -1.09
    notated
    -1.08
    POSITIVE LOGITS
     each
    1.21
     is
    1.20
     außerdem
    1.19
     have
    1.08
     every
    1.07
     لكل
    1.02
     folgenden
    1.00
     각
    0.98
     absence
    0.97
    g
    0.95
    Activations Density 0.021%

    No Known Activations