© 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. 22-GEMMASCOPE-RES-131K
    7. 64918
    Prev
    Next
    INDEX
    Explanations

    first or only instance

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

    The neuron detects ordinal “first” constructions, especially phrases like “first … in …” that denote a pioneering or top-ranked item.

    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
    ׳
    -1.22
    neous
    -1.16
    gering
    -1.12
    ⍽
    -1.08
    editary
    -1.08
    fillment
    -1.07
     variety
    -1.06
     crecimiento
    -1.05
     logistique
    -1.05
    dings
    -1.04
    POSITIVE LOGITS
     in
    1.66
     this
    1.42
     böyle
    1.19
     intellectuelle
    1.11
     подоб
    1.00
     worldwide
    0.99
     unique
    0.97
     Kommune
    0.96
     Mesin
    0.96
     Warga
    0.93
    Activations Density 0.049%

    No Known Activations