© 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. 104480
    Prev
    Next
    INDEX
    Explanations

    following words like 'should', 'english', 'accept', 'thank'

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

    The neuron is strongly activated by sentence-initial (capitalized) words, effectively spotting the beginnings of sentences.

    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
     intéress
    -1.18
    ensión
    -1.17
     ungefähr
    -1.17
     écri
    -1.14
     сначала
    -1.13
    没什么
    -1.11
     trebui
    -1.11
     normalerweise
    -1.09
     przede
    -1.09
     bekan
    -1.08
    POSITIVE LOGITS
     also
    2.75
     therefore
    1.48
     later
    1.48
     then
    1.38
     также
    1.24
     thus
    1.23
    Also
    1.20
     Also
    1.14
     though
    1.13
     will
    1.12
    Activations Density 0.013%

    No Known Activations