© 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. 68343
    Prev
    Next
    INDEX
    Explanations

    first occurrence of things

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

    This neuron activates on tokens that are part of named entities—proper nouns like organization names, team names, labels, conferences, and other capitalized titles.

    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
    に戻
    -0.96
     risposte
    -0.93
     снова
    -0.93
    ština
    -0.93
    Normally
    -0.92
    жья
    -0.89
     polícia
    -0.86
     продол
    -0.86
     Normally
    -0.86
     希望
    -0.85
    POSITIVE LOGITS
     pioneer
    1.56
     early
    1.55
     earliest
    1.36
     damals
    1.34
     first
    1.30
     pioneering
    1.28
     pioneers
    1.23
    初めて
    1.22
    初代
    1.20
     inaugural
    1.20
    Activations Density 0.234%

    No Known Activations