© 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. 22-GEMMASCOPE-RES-131K
    4. 75150
    Prev
    Next
    INDEX
    Explanations

    URLs

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

    The neuron fires on URL components (e.g. “http”, “://”, “/”) indicating it’s detecting the presence of web addresses.

    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
     to
    -1.12
     they
    -0.78
          
    -0.78
    +}\
    -0.75
         
    -0.73
     on
    -0.72
     being
    -0.72
    .(
    -0.72
    これを
    -0.71
     no
    -0.71
    POSITIVE LOGITS
     заинтере
    0.99
     Wohnungen
    0.98
     nossa
    0.94
    ITATION
    0.91
    хьтан
    0.89
    ૮
    0.88
    etje
    0.88
    kswagen
    0.88
    𝓦
    0.87
     binnen
    0.87
    Activations Density 0.002%

    No Known Activations