© 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. 72826
    Prev
    Next
    INDEX
    Explanations

    A followed by specific noun

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

    The neuron activates on isolated uppercase “A” tokens used as answer labels or section headings (e.g. the “A” marking an answer or answer‐score annotation).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    uppercase “A” at the start of a word or as a standalone token, often at the beginning of a line or sentence.

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the single capital letter "A" when it appears as a standalone token or label in technical/programming contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered 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
     diarios
    -2.03
    刳
    -2.03
     și
    -2.02
    黝
    -1.95
     และ
    -1.85
     produz
    -1.80
     şi
    -1.79
    だけです
    -1.77
     anc
    -1.68
    いえば
    -1.66
    POSITIVE LOGITS
     have
    1.91
    也不會
    1.83
    穩
    1.78
    
    1.73
    並不是
    1.73
    6
    1.72
     There
    1.71
     Saltar
    1.67
     from
    1.66
    渽
    1.66
    Activations Density 0.005%

    No Known Activations