© 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. 39539
    Prev
    Next
    INDEX
    Explanations

    listing things with prepositions

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

    This neuron activates on numeric tokens (Arabic numerals, including integers and decimals) that denote quantities.

    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
     principalmente
    -1.16
    ######
    -1.12
     thuy
    -1.09
     arrêter
    -1.08
     remercier
    -1.07
     investissements
    -1.05
     competente
    -1.04
    ended
    -1.03
     emporter
    -1.03
    atl
    -1.02
    POSITIVE LOGITS
     for
    1.38
    řevě
    1.34
     with
    1.34
     after
    1.30
     from
    1.28
    Almost
    1.27
     Before
    1.27
    СЫ
    1.20
     overtly
    1.18
     srd
    1.17
    Activations Density 0.109%

    No Known Activations