© 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. 10-GEMMASCOPE-RES-131K
    7. 61065
    Prev
    Next
    INDEX
    Explanations

    long been

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

    The neuron is specialized to detect the phrase “long been,” i.e. that two-word collocation indicating something has existed or been true for a long time.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    </strong>
    -2.02
     모든
    -1.81
    обои
    -1.73
     gewe
    -1.71
    を使用した
    -1.70
    絝
    -1.70
    珵
    -1.68
    
    -1.66
    {
    -1.65
    hamilan
    -1.63
    POSITIVE LOGITS
    '
    2.44
    c
    2.14
    ll
    2.14
    In
    2.09
    an
    2.08
    в
    2.05
    ly
    1.98
    It
    1.98
    ts
    1.95
    man
    1.95
    Activations Density 0.009%

    No Known Activations