© 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. 10-GEMMASCOPE-RES-131K
    7. 75190
    Prev
    Next
    INDEX
    Explanations

    superlative qualities

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

    The neuron detects superlative modifiers indicating the highest degree or ranking (e.g. “most,” “best,” “furriest”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     
    -1.41
    一樣
    -1.28
    se
    -1.23
     аналоги
    -1.21
    idelberg
    -1.21
     enorme
    -1.16
    在那里
    -1.16
     cortesía
    -1.16
     pasos
    -1.14
    tabria
    -1.14
    POSITIVE LOGITS
     of
    2.33
     sendiri
    1.56
     they
    1.49
     ataupun
    1.44
     However
    1.44
     on
    1.33
     Neither
    1.30
     potuto
    1.30
     Particularly
    1.29
     Ultimately
    1.28
    Activations Density 0.039%

    No Known Activations