© 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. 34-GEMMASCOPE-RES-131K
    7. 103345
    Prev
    Next
    INDEX
    Explanations

    on occasions

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

    The neuron activates on words that serve as noun modifiers—common adjectives and determiners (e.g. every, some, any, gray, drizzly, few).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     على
    -1.21
     on
    -1.16
     на
    -1.13
     روی
    -1.02
     trên
    -0.90
    richtung
    -0.89
     på
    -0.88
    勇者
    -0.88
     Kategorie
    -0.85
     woods
    -0.84
    POSITIVE LOGITS
     occasions
    1.98
     behalf
    1.96
     occasion
    1.70
     ועל
    1.51
     basis
    1.50
     campuses
    1.34
    basis
    1.22
     pretext
    1.21
     grounds
    1.18
    occasion
    1.09
    Activations Density 0.195%

    No Known Activations