© 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. 34-GEMMASCOPE-RES-131K
    7. 107541
    Prev
    Next
    INDEX
    Explanations

    achieve excellent performance

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

    The neuron activates on strong evaluative result–reporting words (like “achieve,” “demonstrate,” “show,” “reveal,” etc.) that announce performance or empirical outcomes in scientific abstracts.

    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
     Whittier
    -0.91
    licates
    -0.83
     soldado
    -0.81
    önes
    -0.80
    にしました
    -0.78
    찼
    -0.77
     condições
    -0.77
     playas
    -0.76
     sofá
    -0.76
     Análisis
    -0.76
    POSITIVE LOGITS
     superior
    1.76
     competitive
    1.55
     outperform
    1.41
     performance
    1.28
    perform
    1.25
    superior
    1.24
     improvements
    1.17
    Superior
    1.14
     Superior
    1.13
     improvement
    1.11
    Activations Density 0.054%

    No Known Activations