© 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. 13392
    Prev
    Next
    INDEX
    Explanations

    maybe youperhaps

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

    The neuron fires on hedge/qualifier adverbs (e.g. “Maybe,” “Perhaps”) that introduce speculation or uncertainty.

    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
     and
    -2.06
    )=>{
    -1.46
    itação
    -1.45
     וכן
    -1.45
    ScopeManager
    -1.42
    etration
    -1.39
    eração
    -1.39
    DebuggerNonUser
    -1.37
     &&
    -1.36
     PUSTAKA
    -1.35
    POSITIVE LOGITS
     they
    3.61
     it
    3.47
     there
    2.97
     we
    2.72
     you
    2.34
     this
    1.86
    ,
    1.71
     she
    1.61
     someone
    1.56
    それでは
    1.52
    Activations Density 0.017%

    No Known Activations