© 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. 13504
    Prev
    Next
    INDEX
    Explanations

    states of being or conditions

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

    The neuron fires on promotional or marketing-style adjectives and phrases (e.g. “enhanced,” “high-quality,” “expanded”) that emphasize product features.

    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
     as
    -3.25
     in
    -3.19
     on
    -3.05
     because
    -3.00
     one
    -2.98
     epitom
    -2.95
     for
    -2.92
     while
    -2.89
     after
    -2.86
     using
    -2.84
    POSITIVE LOGITS
     Dinas
    2.72
     !!!!
    2.50
     revered
    2.44
     goofy
    2.44
     ridiculously
    2.42
     AMAZING
    2.41
     unbelievably
    2.41
     delightfully
    2.39
     ſever
    2.39
     trám
    2.38
    Activations Density 0.189%

    No Known Activations