© 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. 22-GEMMASCOPE-RES-131K
    7. 13058
    Prev
    Next
    INDEX
    Explanations

    highly positive adjectives

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

    The neuron detects non-standard, high-Unicode (often private-use or PDF-artifact) glyphs—i.e. unusual characters outside ordinary ASCII.

    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
     this
    -3.80
     four
    -3.78
     undeniably
    -3.77
     any
    -3.73
     cleverly
    -3.72
     at
    -3.69
     on
    -3.64
     what
    -3.61
     three
    -3.61
     putting
    -3.50
    POSITIVE LOGITS
     delightfully
    3.80
     luscious
    3.80
    
    3.78
     strikingly
    3.66
     insanely
    3.59
     ridiculously
    3.58
     captivating
    3.55
     acclaimed
    3.53
     unbelievably
    3.52
     deliciously
    3.47
    Activations Density 0.013%

    No Known Activations