© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 71959
    Prev
    Next
    INDEX
    Explanations

    searching for information

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

    The neuron responds to words and phrases that indicate doing an online search or looking something up (e.g. “search,” “google(d),” “googling,” “go through”).

    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
     cuando
    -1.20
     when
    -1.20
    ↵↵↵
    -1.11
    cuando
    -1.10
    ↵↵↵↵↵↵↵
    -1.07
    ↵↵↵↵↵
    -1.06
    ↵↵↵↵↵↵
    -1.04
    ↵↵↵↵
    -1.02
     is
    -1.02
    ↵↵↵↵↵↵↵↵↵
    -1.02
    POSITIVE LOGITS
     search
    1.23
     keyword
    1.15
     keywords
    1.10
    search
    1.09
    検索
    1.03
     blanches
    1.03
     búsqueda
    1.03
     lançado
    1.02
    
    1.01
     into
    1.00
    Activations Density 0.069%

    No Known Activations