© 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. 34-GEMMASCOPE-RES-131K
    7. 2893
    Prev
    Next
    INDEX
    Explanations

    that's why

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

    The neuron consistently fires on occurrences of the copular “is” (and its contraction ’s), especially in constructions like “that is,” “which is,” or “that’s.”

    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
     how
    -1.38
     what
    -1.17
    どう
    -0.98
    How
    -0.96
     raisons
    -0.89
     cómo
    -0.85
     hvordan
    -0.85
    how
    -0.83
     rôles
    -0.82
    バージョン
    -0.82
    POSITIVE LOGITS
     part
    1.32
     one
    1.16
     поэтому
    1.14
    latego
    1.12
     Поэтому
    1.10
     why
    1.05
     yksi
    1.03
    Fortunately
    1.01
     część
    0.99
     precisely
    0.97
    Activations Density 0.020%

    No Known Activations