© 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. 9843
    Prev
    Next
    INDEX
    Explanations

    claims about groups or theories

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

    This neuron lights up on relatively uncommon, content-bearing words (i.e. low-frequency or “rare” tokens) rather than on common function 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_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
    Gilla
    -1.09
    '''
    -1.04
     Konflikt
    -1.02
     was
    -1.00
    -1.00
     fortfarande
    -0.99
     his
    -0.98
    ##
    -0.97
    ></
    -0.97
     ändå
    -0.96
    POSITIVE LOGITS
     ২০১
    1.30
     ideen
    1.27
     often
    1.17
     nadru
    1.13
    Recently
    1.13
    Various
    1.13
    комнат
    1.13
     nader
    1.12
    Often
    1.09
     kampen
    1.09
    Activations Density 0.106%

    No Known Activations