© 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. 98427
    Prev
    Next
    INDEX
    Explanations

    foreign language fragments

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

    This neuron fires on isolated numeric tokens or sequences of uppercase letters (e.g. section numbers, acronyms, or all‐caps headings).

    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
     nejsou
    -1.55
     ſont
    -1.52
     strikingly
    -1.44
     soooo
    -1.40
     breathtaking
    -1.39
     sooo
    -1.38
    seguir
    -1.38
    probably
    -1.38
     satisfactor
    -1.37
    wanna
    -1.37
    POSITIVE LOGITS
     one
    1.95
     that
    1.63
     time
    1.61
     continual
    1.51
     increase
    1.49
     many
    1.39
     method
    1.36
     system
    1.34
     you
    1.33
     first
    1.33
    Activations Density 0.011%

    No Known Activations