© 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. 64619
    Prev
    Next
    INDEX
    Explanations

    but, that, though, whether

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

    The neuron is strongly activated by discourse-linking words—especially subordinating or contrastive conjunctions like “but,” “that,” and “when.”

    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
     or
    -2.56
     at
    -2.48
     as
    -2.45
     using
    -2.41
     by
    -2.39
     ostensibly
    -2.38
     in
    -2.38
     so
    -2.36
     from
    -2.36
     only
    -2.31
    POSITIVE LOGITS
     Dinas
    2.33
    commonly
    2.14
     😁
    2.03
     .......
    2.03
     majest
    2.00
    
    1.99
     garan
    1.98
     poft
    1.98
    킁
    1.97
    
    1.96
    Activations Density 0.114%

    No Known Activations