© 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. 22-GEMMASCOPE-RES-131K
    7. 67709
    Prev
    Next
    INDEX
    Explanations

    I followed by common verbs or contractions

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

    The neuron is picking out subject personal pronouns (especially “I” and similar pronoun 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_22/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
    кас
    -1.05
    可能是
    -1.04
     тот
    -1.04
    もあって
    -1.03
    lity
    -1.03
     Grüßen
    -1.03
    促進
    -0.99
    齟
    -0.99
     APPROVED
    -0.99
     توضی
    -0.98
    POSITIVE LOGITS
     so
    1.21
     which
    1.16
     as
    1.12
     if
    1.06
     several
    1.03
     fierce
    1.03
     exquisitely
    1.02
     and
    1.01
     unparalleled
    1.01
     supremely
    0.99
    Activations Density 0.024%

    No Known Activations