© 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. 10976
    Prev
    Next
    INDEX
    Explanations

    Africa:language codes, Q:, license headers

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

    The neuron is detecting double‐quote characters—i.e. the quotation delimiters that mark the start or end of string literals.

    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
    zdo
    -1.24
    なのかな
    -1.23
    ITCHEN
    -1.10
    -1.08
     fortaleza
    -1.08
     stör
    -1.07
     blusa
    -1.06
     dátum
    -1.06
     respirar
    -1.05
    tersebut
    -1.05
    POSITIVE LOGITS
     чаще
    0.97
    lysm
    0.95
     genoc
    0.94
    TLDR
    0.94
    甚至是
    0.92
    free
    0.92
    ↵↵
    0.91
    k
    0.91
     jagged
    0.91
     सिद्ध
    0.90
    Activations Density 0.000%

    No Known Activations