© 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. 33524
    Prev
    Next
    INDEX
    Explanations

    but and

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

    The neuron strongly activates on the adversative conjunction “but,” i.e. when a sentence shift or contrast is introduced.

    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
     tempel
    -1.33
     spesifik
    -1.22
     gril
    -1.14
     ekst
    -1.13
     palme
    -1.13
     musli
    -1.12
     graden
    -1.09
     tille
    -1.08
     parati
    -1.06
     Portug
    -1.05
    POSITIVE LOGITS
     باعث
    1.66
     allow
    1.51
    ทำให้
    1.48
     cause
    1.37
     enable
    1.37
     khiến
    1.35
     help
    1.32
     also
    1.31
     it
    1.27
     provide
    1.16
    Activations Density 0.113%

    No Known Activations