© 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. 10640
    Prev
    Next
    INDEX
    Explanations

    I and we pronoun usage

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

    The neuron responds to first-person pronouns, especially “I” and “we,” marking self-references in the text.

    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
    枇
    -1.61
    Fprintf
    -1.57
     on
    -1.53
     pamoja
    -1.52
    这让
    -1.45
     nagu
    -1.41
    listFiles
    -1.40
    (;;)
    -1.39
    ！！！！
    -1.38
    ############
    -1.36
    POSITIVE LOGITS
     who
    1.58
     the
    1.51
     neither
    1.49
     siguran
    1.48
    ろん
    1.48
     deberán
    1.46
     ensured
    1.43
    Belgique
    1.42
     eigentliche
    1.42
     needed
    1.41
    Activations Density 0.198%

    No Known Activations