© 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. 75058
    Prev
    Next
    INDEX
    Explanations

    application no

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

    The neuron flags boilerplate or structured‐metadata tokens typical of formal legal, regulatory, or patent documents (e.g. “rule,” “Patent Application No.,” “priority,” “Confidentiality Notice,” 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
     Volkes
    -1.16
     tunik
    -1.16
    ooooooo
    -1.16
     akku
    -1.13
     nî
    -1.12
    ๗
    -1.12
     dekol
    -1.09
     kerugian
    -1.08
    hộp
    -1.07
    zeń
    -1.06
    POSITIVE LOGITS
     infallible
    0.99
    DIMENSION
    0.98
    也就是说
    0.96
    优秀的
    0.93
     своим
    0.93
     appreciable
    0.91
    明显的
    0.91
    顺便
    0.91
     courants
    0.89
    Frankly
    0.89
    Activations Density 0.094%

    No Known Activations