© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 39077
    Prev
    Next
    INDEX
    Explanations

    as comparative

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

    The neuron activates on comparative “as … as” constructions (and close variants like “nearly as,” “quite as,” or “equally as”) linking an adjective or adverb on both sides.

    oai_token-act-pair · o4-miniTriggered 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
     strikingly
    -2.17
     következő
    -2.06
     conspicuously
    -2.03
    ⸸
    -2.00
    でも
    -2.00
     vastly
    -1.99
     menacing
    -1.98
    ⃨
    -1.90
    mpagne
    -1.88
    骉
    -1.85
    POSITIVE LOGITS
    v
    1.80
    mo
    1.78
    f
    1.73
     that
    1.72
     or
    1.69
    z
    1.64
    to
    1.62
    ekš
    1.60
    t
    1.60
     from
    1.58
    Activations Density 0.004%

    No Known Activations