© 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. 34-GEMMASCOPE-RES-131K
    4. 50468
    Prev
    Next
    INDEX
    Explanations

    expressing confusion or difficulty

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

    This neuron fires on first-person statements—especially occurrences of “I” (and related forms like I’m, my, feel, like, etc.) indicating the speaker referring to themselves.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     チラ
    -1.07
     kug
    -1.03
     ヌード
    -0.96
     yaka
    -0.95
    ņu
    -0.94
    铆
    -0.93
    しかし
    -0.93
     vay
    -0.93
    ersatz
    -0.93
     pérd
    -0.92
    POSITIVE LOGITS
     actitudes
    1.19
     scary
    1.05
     bad
    1.05
     weird
    1.05
     influenced
    1.02
     termed
    1.01
     complicated
    1.01
     aggressive
    0.98
     stress
    0.96
     curving
    0.94
    Activations Density 0.085%

    No Known Activations