© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 79817
    Prev
    Next
    INDEX
    Explanations

    contrast or but

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

    The neuron strongly activates on the adversative conjunction “but” (and similar contrast markers).

    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
     also
    -0.89
    ėse
    -0.86
     modelAndView
    -0.86
     then
    -0.85
    MatButtonModule
    -0.84
     alors
    -0.84
     大会
    -0.83
    有多
    -0.81
     già
    -0.80
    了好
    -0.80
    POSITIVE LOGITS
    这一次
    1.05
     apparti
    0.99
    这次
    0.93
     surpasses
    0.93
    ไหน
    0.93
     nutri
    0.92
    什么意思
    0.92
    というもの
    0.91
    ressee
    0.90
    是最
    0.89
    Activations Density 0.059%

    No Known Activations