© 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. 34-GEMMASCOPE-RES-131K
    7. 99431
    Prev
    Next
    INDEX
    Explanations

    as if / as though clauses

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

    This neuron strongly activates on short, high-frequency function words (especially pronouns and linking verbs like “he,” “she,” “it,” “was,” “they,” “as,” “if,” 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_34/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
     seeming
    -1.12
     seem
    -1.08
    ά
    -1.04
    seem
    -1.02
     sembler
    -1.01
     úgy
    -1.00
    λέον
    -0.98
     يبدو
    -0.96
     parece
    -0.96
    貌
    -0.95
    POSITIVE LOGITS
     were
    1.06
     audition
    1.00
     someone
    0.93
     sedang
    0.91
     belongs
    0.91
    最初から
    0.90
     suddenly
    0.89
     algun
    0.88
     innumerable
    0.88
     just
    0.88
    Activations Density 0.034%

    No Known Activations