© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 130991
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on personal, first-person narrative language—e.g. “I,” “my,” “husband,” and other subjective commentary indicating the speaker’s own experience.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    film reviews or titles

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    fucking
    -1.05
    melons
    -1.04
     goddamn
    -1.00
     获得
    -0.98
    omorphisms
    -0.94
     hotell
    -0.94
     joyous
    -0.93
    叒
    -0.93
    thodoxy
    -0.93
    😊
    -0.93
    POSITIVE LOGITS
     this
    1.45
     to
    1.16
    andangan
    1.14
     would
    1.09
     or
    1.06
     should
    1.05
     disappointing
    1.04
     disappointed
    1.00
     seemed
    0.98
     aurait
    0.98
    Activations Density 0.008%

    No Known Activations