© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 27450
    Prev
    Next
    INDEX
    Explanations

    proper nouns in various scripts

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

    The neuron fires on single‐letter or short multi‐letter tokens that are uppercase or title-case initials (i.e. the leading capital letters of names, titles, language codes, etc.).

    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
     I
    -1.39
    itoneal
    -1.30
     bolí
    -1.27
     artesan
    -1.24
     kiệm
    -1.23
     biela
    -1.23
    ∛
    -1.23
    鱂
    -1.22
     manualidades
    -1.21
    =>$
    -1.20
    POSITIVE LOGITS
    0
    1.36
    O
    1.35
    B
    1.34
    N
    1.27
    '
    1.27
    7
    1.23
    3
    1.22
    6
    1.20
    5
    1.19
    8
    1.15
    Activations Density 0.051%

    No Known Activations