© 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. 42300
    Prev
    Next
    INDEX
    Explanations

    speaking in first person

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

    The neuron detects tokens where the author refers to themselves in the first person (e.g. “I,” “I’ve,” “I’m,” etc.).

    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
     is
    -2.61
    </em>
    -2.59
     of
    -2.45
     has
    -2.41
     '
    -2.33
     "
    -2.13
     (
    -2.03
    </h2>
    -2.03
     had
    -2.02
     then
    -1.97
    POSITIVE LOGITS
     ἀ
    2.16
    睒
    2.08
     Reverso
    1.99
    獁
    1.99
    ‘‘
    1.97
    ፑ
    1.96
     particules
    1.94
    
    1.93
     bors
    1.91
    瞜
    1.91
    Activations Density 0.023%

    No Known Activations