© 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. 35136
    Prev
    Next
    INDEX
    Explanations

    multi-language text

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

    This neuron fires on phrasing where the writer asks for the meaning or clarification of a specific term (e.g. “What do you mean by the term ‘synced to the WiFi signal’?”).

    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
    Vastaa
    -1.05
    ǎ
    -0.98
     POETRY
    -0.98
    ᗨ
    -0.97
    podob
    -0.94
    ėjas
    -0.94
     trampa
    -0.94
     calavera
    -0.90
     Universitas
    -0.90
     Богу
    -0.89
    POSITIVE LOGITS
     unless
    1.02
    ਜ
    0.90
     described
    0.90
     espíritu
    0.89
    这么
    0.89
     Unless
    0.89
     are
    0.87
     such
    0.87
    だけで
    0.86
     use
    0.86
    Activations Density 0.035%

    No Known Activations