© 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. 49185
    Prev
    Next
    INDEX
    Explanations

    compound noun phrases

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

    The neuron fires on the first word acting as a modifier in two-word noun compounds (e.g. “dining” in “dining experience,” “campaign” in “campaign trail,” “diffraction” in “diffraction limit,” 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
    凄く
    -1.21
    すっ
    -1.20
    だいぶ
    -1.18
     situações
    -1.16
     málaga
    -1.09
    凄い
    -1.09
    うわ
    -1.08
     horrifying
    -1.08
    かわい
    -1.05
    proyecto
    -1.04
    POSITIVE LOGITS
    when
    1.28
     after
    1.20
     when
    1.19
    著名的
    1.09
    esss
    1.02
    特别是
    1.02
    🎃
    0.99
    ﾁ
    0.94
     öppna
    0.94
    了许多
    0.93
    Activations Density 0.283%

    No Known Activations