© 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. 99405
    Prev
    Next
    INDEX
    Explanations

    than yesterday's

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

    The neuron is highly attuned to words that signal comparative or contrasting contexts (e.g. “compared,” “from,” “common,” “ever”).

    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
    之后
    -1.04
     subsequent
    -1.00
    later
    -0.98
    中文
    -0.95
    urant
    -0.95
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    -0.90
     späteren
    -0.89
     ò
    -0.89
     redan
    -0.88
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    -0.88
    POSITIVE LOGITS
     ten
    2.05
     twenty
    2.03
     five
    1.96
     thirty
    1.88
     yesterday
    1.86
     few
    1.84
     before
    1.78
     fifteen
    1.69
     vroeger
    1.66
    前は
    1.59
    Activations Density 0.054%

    No Known Activations