© 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. 15719
    Prev
    Next
    INDEX
    Explanations

    but / however

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

    This neuron detects discourse‐level conjunctions and adverbs (e.g. “but,” “however,” “as,” “also”) that signal contrast or continuation.

    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
    所以在
    -1.05
    を
    -1.01
     [],
    -0.98
     turist
    -0.95
    sssss
    -0.93
    ssss
    -0.91
    нейших
    -0.87
    ñores
    -0.86
     waterfall
    -0.86
     kapital
    -0.85
    POSITIVE LOGITS
     oczywiście
    1.23
    来週
    1.18
    久し
    1.16
    為に
    1.16
     toutefois
    1.15
    ごめんなさい
    1.14
    事も
    1.12
    明日は
    1.11
    んですよ
    1.10
    ようになって
    1.10
    Activations Density 0.059%

    No Known Activations