© 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. 10-GEMMASCOPE-RES-131K
    7. 66665
    Prev
    Next
    INDEX
    Explanations

    as part of

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

    The neuron specializes in detecting the word “part,” especially when it appears in the phrase “as part of.”

    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
     same
    -2.16
     use
    -1.74
     only
    -1.68
     different
    -1.61
    =
    -1.52
     both
    -1.52
    J
    -1.50
     various
    -1.49
    >
    -1.48
     those
    -1.48
    POSITIVE LOGITS
     of
    1.63
    
    1.54
     Setelah
    1.49
    𓐍
    1.49
     Sehingga
    1.44
     餐厅
    1.41
     Kemudian
    1.39
    PropertyChanging
    1.38
     Berikut
    1.37
     破解
    1.37
    Activations Density 0.042%

    No Known Activations