© 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. 74686
    Prev
    Next
    INDEX
    Explanations

    word beginnings

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

    The neuron is primarily firing on personal pronouns and their informal contraction forms (e.g. “I,” “theyre,” “youre”) as tokens.

    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
    ,"
    -2.84
    $
    -2.48
    }\
    -2.31
    If
    -2.30
    ）
    -2.17
    𔘓
    -2.06
     celebrates
    -2.03
    --"
    -2.03
    When
    -2.02
     姉
    -2.02
    POSITIVE LOGITS
     It
    3.00
     a
    2.28
     “(
    1.98
    更高的
    1.95
     😭
    1.95
    になっている
    1.94
     “…
    1.94
     conhecida
    1.93
    ization
    1.90
     tomó
    1.90
    Activations Density 0.000%

    No Known Activations