© 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. 15939
    Prev
    Next
    INDEX
    Explanations

    usage of although

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

    The neuron consistently activates on the concessive linking word “Although,” i.e. the discourse marker introducing a contrast clause.

    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
     and
    -2.95
    菝
    -2.27
    鏷
    -2.25
    瀲
    -2.14
     เรา
    -1.93
    </b>
    -1.91
     เพราะ
    -1.91
    퐹
    -1.88
    銩
    -1.86
    Tó
    -1.84
    POSITIVE LOGITS
     it
    2.86
    {
    2.77
     we
    2.41
    The
    2.34
    .
    2.27
     the
    2.22
     ersche
    2.20
    '
    2.11
     behaup
    2.06
     they
    2.03
    Activations Density 0.016%

    No Known Activations