© 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. 34-GEMMASCOPE-RES-131K
    7. 81494
    Prev
    Next
    INDEX
    Explanations

    goes, went, going, gone

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

    The main thing this neuron does is detect occurrences of the verb “go” in any form (e.g. go, goes, going – and even its German equivalent “geht”).

    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
     provved
    -0.98
     contatto
    -0.85
     reflek
    -0.77
    ▵
    -0.77
     ony
    -0.77
     regola
    -0.77
     memas
    -0.77
    ⋀
    -0.75
    וּ
    -0.75
     noth
    -0.74
    POSITIVE LOGITS
     into
    1.21
     verloren
    0.92
     through
    0.83
     off
    0.82
    kul
    0.79
     attrition
    0.78
    čios
    0.78
     neki
    0.77
     apart
    0.77
     away
    0.76
    Activations Density 0.007%

    No Known Activations