© 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. 39388
    Prev
    Next
    INDEX
    Explanations

    unsigned short type

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

    This neuron detects tokens that begin new structural blocks or sections (e.g. the very first content word of a paragraph, code block, or heading).

    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
     its
    -1.45
     more
    -1.19
     Komun
    -1.13
     to
    -1.10
    …！
    -1.08
     ū
    -1.07
     indeed
    -1.04
     wirklich
    -1.02
     barely
    -0.98
    ردد
    -0.98
    POSITIVE LOGITS
     azonban
    1.20
     imidlertid
    1.14
     ieder
    1.13
     eenvou
    1.10
     moeili
    1.09
    COVID
    1.09
    will
    1.09
    2
    1.05
     vigencia
    1.05
     schrij
    1.05
    Activations Density 0.001%

    No Known Activations