© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100401
    Prev
    Next
    INDEX
    Explanations

    repeated intensifiers

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

    The neuron fires on emphatic repeated intensifiers—instances of “very, very” (often with the following comma).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     inconvénients
    -2.06
    ⦊
    -2.05
    "¿
    -1.96
     rigue
    -1.94
    之际
    -1.91
    atguigu
    -1.85
     vanwege
    -1.82
    赑
    -1.80
     レスト
    -1.79
    珖
    -1.79
    POSITIVE LOGITS
    mybatisplus
    1.87
    ті
    1.84
    </th>
    1.80
    しました
    1.77
    </sub>
    1.77
     *
    1.74
    DockStyle
    1.73
    ꔹ
    1.72
     />
    1.69
    ">(</
    1.67
    Activations Density 0.009%

    No Known Activations