© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 68678
    Prev
    Next
    INDEX
    Explanations

    initials and abbreviations

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

    The neuron activates on isolated single-letter uppercase tokens (e.g. author initials or similar one-letter capitals).

    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
     what
    -1.98
     all
    -1.94
     that
    -1.88
     either
    -1.85
     first
    -1.84
     necessary
    -1.64
     these
    -1.61
     how
    -1.60
     as
    -1.55
     just
    -1.50
    POSITIVE LOGITS
     brillant
    1.75
     refroid
    1.71
     rafra
    1.66
    ⬈
    1.63
    蚰
    1.58
     ambiti
    1.55
     manquer
    1.52
     diriger
    1.52
    ܆
    1.52
     solides
    1.52
    Activations Density 0.006%

    No Known Activations