© 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. 28346
    Prev
    Next
    INDEX
    Explanations

    descriptors forming compound nouns

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

    The neuron is triggered by technical or domain-specific descriptive modifiers—multi-syllable adjectives and compound qualifiers that precede nouns in formal and scientific writing.

    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
     aident
    -1.31
    立马
    -1.20
     rendent
    -1.16
     enorme
    -1.15
    particularly
    -1.12
     ۱۵
    -1.11
     masterful
    -1.10
     famed
    -1.09
     absurd
    -1.09
     crucial
    -1.09
    POSITIVE LOGITS
     it
    1.45
     through
    1.32
     Those
    1.28
    During
    1.27
     like
    1.26
     bringing
    1.23
     with
    1.22
     nestled
    1.22
     diminish
    1.20
     less
    1.20
    Activations Density 0.131%

    No Known Activations