© 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. 102155
    Prev
    Next
    INDEX
    Explanations

    code, braces, and tags

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

    the pipe character (|) used as a delimiter or separator in code and data formats.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron detects literal pipe characters (the “|” delimiter) in code or regex contexts.

    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
     mostrando
    -1.13
    crimin
    -1.05
     residuos
    -1.02
     女神
    -1.01
    nyard
    -1.00
     alimentación
    -0.99
     羊
    -0.99
    hever
    -0.99
     Perfon
    -0.98
     完了
    -0.98
    POSITIVE LOGITS
     запах
    1.20
     Dernière
    1.20
     behaves
    1.19
    avras
    1.18
    bondale
    1.16
     Whilst
    1.13
     While
    1.11
     November
    1.09
    imentos
    1.07
    OWN
    1.06
    Activations Density 0.099%

    No Known Activations