© 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. 49156
    Prev
    Next
    INDEX
    Explanations

    prepositions for location and relation

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

    The neuron strongly activates on English prepositions (short function words like “in,” “on,” “with,” “to,” “for,” “under,” “upon,” etc.).

    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
    将其
    -1.06
    ־ה
    -1.00
     рекомендую
    -0.93
     большин
    -0.91
    •
    -0.90
     зрения
    -0.89
     初始化
    -0.89
     wailing
    -0.87
    つなが
    -0.86
    辄
    -0.85
    POSITIVE LOGITS
     the
    2.08
     every
    1.50
     your
    1.40
     our
    1.28
     their
    1.15
     pomy
    1.10
     another
    1.10
     those
    1.06
    spaceBetween
    1.02
     each
    0.98
    Activations Density 0.044%

    No Known Activations