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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 14519
    Prev
    Next
    INDEX
    Explanations

    game titles and proper nouns

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

    The neuron strongly responds to high-information content words—especially titles, proper names, and prominent nouns/verbs that signal the main topics or entities of an article.

    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
     often
    -1.28
     when
    -1.23
     suoi
    -1.19
     sometimes
    -1.17
     L
    -1.16
     really
    -1.15
     until
    -1.15
     S
    -1.14
     T
    -1.13
     manchmal
    -1.13
    POSITIVE LOGITS
     serão
    1.44
     quadrinhos
    1.35
     جدید
    1.34
     signifikan
    1.30
     kompati
    1.30
    据
    1.29
    änge
    1.28
    presumably
    1.28
     будет
    1.27
     saveurs
    1.27
    Activations Density 0.041%

    No Known Activations