© 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. 39987
    Prev
    Next
    INDEX
    Explanations

    of the

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

    The neuron flags definitional linking phrases—especially occurrences of “form of” (and similar “X is a form of Y”) indicating a definition.

    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
    )))))
    -1.09
     and
    -1.06
     utilizando
    -1.06
     noen
    -1.02
    gee
    -1.00
     очеред
    -0.99
    chka
    -0.98
    trage
    -0.96
     ※
    -0.96
    🤘
    -0.95
    POSITIVE LOGITS
    Année
    1.14
     gestalten
    1.13
     zakresie
    1.13
    انيا
    1.13
     démoc
    1.13
    entreprises
    1.12
     nerfs
    1.09
    moteur
    1.08
    というわけで
    1.07
     духовке
    1.07
    Activations Density 0.108%

    No Known Activations