© 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. 38339
    Prev
    Next
    INDEX
    Explanations

    greeting and hello

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

    The neuron strongly activates on common English words—particularly function words, pronouns, verbs, and other high‐frequency English vocabulary—while ignoring non‐English or zero‐activation content.

    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
    descriptions
    -0.79
    Слу
    -0.78
    pendium
    -0.75
    谏
    -0.74
     lato
    -0.73
    ሚ
    -0.72
    teni
    -0.72
     latitudes
    -0.71
     Commitment
    -0.71
    ख्या
    -0.71
    POSITIVE LOGITS
     greeting
    5.09
     greetings
    4.72
     greet
    4.28
     greeted
    3.73
     greets
    3.70
     Greeting
    3.64
    greeting
    3.64
    Greeting
    3.55
     Greetings
    3.53
     hello
    3.42
    Activations Density 0.075%

    No Known Activations