© 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. 70401
    Prev
    Next
    INDEX
    Explanations

    superlative or extreme states

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

    The neuron detects the "’s" suffix (the possessive or contraction ending) on words.

    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
     through
    -1.99
     because
    -1.94
     during
    -1.88
     and
    -1.77
     while
    -1.77
     their
    -1.73
    ogues
    -1.73
     enhances
    -1.70
     analisis
    -1.70
    neling
    -1.70
    POSITIVE LOGITS
    苤
    1.91
     kinda
    1.86
     REALLY
    1.75
     ridiculously
    1.63
     trám
    1.63
    didn
    1.60
     montrer
    1.60
     yearning
    1.59
     serene
    1.58
    
    1.55
    Activations Density 0.039%

    No Known Activations