© 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. 130556
    Prev
    Next
    INDEX
    Explanations

    programming code elements

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

    The neuron fires on sequences defining or listing character codes—especially numeric and hexadecimal literals (digits, “0x…” forms, “~”, etc.) used to specify allowable characters.

    oai_token-act-pair · o4-miniTriggered 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
     of
    -1.84
     like
    -1.76
     his
    -1.55
     here
    -1.45
     him
    -1.42
     this
    -1.35
     currently
    -1.34
     compared
    -1.31
     life
    -1.29
     up
    -1.28
    POSITIVE LOGITS
     Sudah
    1.34
     bluse
    1.34
    不僅
    1.34
     mengumumkan
    1.31
     aussitôt
    1.29
     скачать
    1.29
    damit
    1.28
    獷
    1.28
     britannique
    1.27
    為什麼
    1.27
    Activations Density 0.321%

    No Known Activations