© 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. 10-GEMMASCOPE-RES-131K
    7. 42207
    Prev
    Next
    INDEX
    Explanations

    personal information or attributes

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

    The neuron is sensitive to semantically rich “content” words—main verbs and nouns carrying the core meaning—rather than to grammatical function words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     nedan
    -1.27
     бюдже
    -1.19
     și
    -1.12
     you
    -1.06
    君は
    -1.04
     thirteen
    -1.02
     Spain
    -1.02
    出来ない
    -1.01
     masih
    -1.00
     размере
    -1.00
    POSITIVE LOGITS
     آنچه
    1.12
     up
    1.05
    ANIE
    1.05
    ortic
    0.99
    */
    0.98
     происходит
    0.98
     取り付け
    0.97
     другой
    0.95
    textura
    0.95
    REFERRED
    0.94
    Activations Density 0.178%

    No Known Activations