© 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. 34-GEMMASCOPE-RES-131K
    7. 59157
    Prev
    Next
    INDEX
    Explanations

    articles followed by nouns

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

    The neuron strongly activates on virtually every content word in a news‐style report about an Israeli military lieutenant warning of “grave danger,” so it’s detecting tokens in that specific conflict‐reporting context.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     tré
    -1.05
    alties
    -1.03
    عتبر
    -0.96
     ofta
    -0.92
     aile
    -0.91
    idable
    -0.91
     skall
    -0.91
     BioLib
    -0.91
    lably
    -0.90
     nú
    -0.88
    POSITIVE LOGITS
     estadounidense
    1.05
     foodstuffs
    0.96
     대해
    0.94
    himp
    0.91
     dear
    0.86
    であ
    0.84
     neoliber
    0.83
     YouTuber
    0.83
     vecchia
    0.83
     satan
    0.83
    Activations Density 0.009%

    No Known Activations