© 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. 22-GEMMASCOPE-RES-131K
    7. 27491
    Prev
    Next
    INDEX
    Explanations

    is named

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

    The neuron fires strongly on passive‐voice constructions—particularly occurrences of “was” (or “were”) combined with a past participle.

    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
    omalies
    -0.78
     لإ
    -0.77
    وأضاف
    -0.77
    थि
    -0.77
    的很
    -0.77
    /**
    -0.76
     parfois
    -0.76
    шом
    -0.74
     europeo
    -0.73
     fiecare
    -0.73
    POSITIVE LOGITS
     named
    2.59
     called
    2.45
     name
    2.45
     bernama
    2.34
     nime
    2.03
     names
    1.99
    名称
    1.94
     renamed
    1.94
     называ
    1.88
    ชื่อ
    1.87
    Activations Density 0.105%

    No Known Activations