© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 1409
    Prev
    Next
    INDEX
    Explanations

    names and specific noun phrases

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

    This neuron fires on the first (often capitalized) word or label beginning a new section, heading, speaker cue, or passage in the text.

    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
     than
    -1.53
     then
    -1.38
     گرف
    -1.34
    必要があります
    -1.28
     vermelhas
    -1.24
     of
    -1.18
    polated
    -1.17
     همان
    -1.16
     穴
    -1.13
    有著
    -1.13
    POSITIVE LOGITS
     empezar
    1.42
    How
    1.32
    Why
    1.26
    tais
    1.22
     Why
    1.21
     ۱۴۰۰
    1.20
     compartir
    1.18
    mú
    1.14
    Some
    1.14
     Besides
    1.13
    Activations Density 0.163%

    No Known Activations