© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 25206
    Prev
    Next
    INDEX
    Explanations

    then clauses

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

    The neuron activates on markers of conditional clauses—particularly the token “if” and its associated commas (and sometimes “then”) signaling an if…then construction.

    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
     dianteiro
    -1.12
     traseiro
    -1.11
    łej
    -1.11
     chapéu
    -1.09
     soprattutto
    -1.09
     レザー
    -1.08
    也在
    -1.04
    ܆
    -1.04
    łych
    -1.03
     frequentemente
    -1.02
    POSITIVE LOGITS
     then
    4.91
    Then
    2.38
     it
    2.31
     Then
    2.30
    then
    2.28
     or
    2.16
     entonces
    2.16
     тогда
    2.08
     thì
    2.06
     akkor
    1.95
    Activations Density 0.063%

    No Known Activations