© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 128348
    Prev
    Next
    INDEX
    Explanations

    functional features

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

    The neuron detects descriptive, feature-highlighting adjectives (e.g. “heated,” “powered,” “light-up,” “turbo-charged,” “electronic”) that modify product attributes.

    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
     reikia
    -1.31
     beobachtet
    -1.21
     jir
    -1.20
    릉
    -1.19
    先日
    -1.16
    MEG
    -1.16
     baseUrl
    -1.13
     verändert
    -1.13
    はある
    -1.12
     dabar
    -1.12
    POSITIVE LOGITS
     основные
    1.40
     maravillas
    1.37
     большие
    1.34
     dueños
    1.20
     appel
    1.20
     эти
    1.19
     все
    1.19
     тыс
    1.18
     jedoch
    1.17
     dieser
    1.16
    Activations Density 0.121%

    No Known Activations