© 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. 19120
    Prev
    Next
    INDEX
    Explanations

    describing or classifying using 'as'

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

    The neuron fires on occurrences of “as” when it’s part of attributive or comparative phrases (e.g. “described as,” “regarded as,” “thought of as”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    uses of “as” that link someone or something to a role, description, or category (attributive/comparative constructions).

    oai_token-act-pair · gpt-5Triggered by @jyhe0408

    the word "as" when it precedes characterizations, descriptions, or classifications of people, things, or concepts.

    oai_token-act-pair · claude-4-5-sonnetTriggered 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
    for
    -2.14
    9
    -2.02
    6
    -1.91
    8
    -1.88
     efforts
    -1.88
    ing
    -1.86
    5
    -1.80
    ]
    -1.73
    4
    -1.70
    In
    -1.66
    POSITIVE LOGITS
     a
    2.38
     an
    1.76
    ⃙
    1.74
    驺
    1.67
     jde
    1.63
    </h1>
    1.61
    mdan
    1.60
     bereiken
    1.60
     sebuah
    1.56
    哢
    1.52
    Activations Density 0.059%

    No Known Activations