© 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. 10-GEMMASCOPE-RES-131K
    7. 47752
    Prev
    Next
    INDEX
    Explanations

    compensation

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

    The main thing this neuron does is detect occurrences of the word “compensation.”

    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
    '
    -2.91
    derdag
    -2.80
     Heilige
    -2.69
    
    -2.64
    关键
    -2.55
    很棒
    -2.47
    很容易
    -2.45
    角度
    -2.41
    re
    -2.41
    end
    -2.39
    POSITIVE LOGITS
    でしたが
    3.34
    慥
    3.20
     parachoque
    2.84
    一说
    2.69
    酲
    2.61
    ↵
    2.56
    を設置
    2.53
    酝
    2.50
    UNTAIN
    2.42
    使って
    2.42
    Activations Density 0.005%

    No Known Activations