© 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. 40050
    Prev
    Next
    INDEX
    Explanations

    probably followed by certain words

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

    The neuron fires on hedging or uncertainty expressions (e.g. “probably,” qualifiers like “more,” and other words signaling doubt or tentative statements).

    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
     parfois
    -1.67
    </h2>
    -1.49
     semblent
    -1.41
     soms
    -1.37
    parently
    -1.32
     renon
    -1.31
     sembl
    -1.30
     pudiera
    -1.28
    طنين
    -1.26
     appara
    -1.25
    POSITIVE LOGITS
     most
    1.59
     will
    1.48
     more
    1.24
     somewhere
    1.22
     won
    1.19
     reinen
    1.18
    2
    1.18
    8
    1.18
     wouldn
    1.17
    4
    1.17
    Activations Density 0.029%

    No Known Activations