© 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. 14003
    Prev
    Next
    INDEX
    Explanations

    fight

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

    This neuron fires on words denoting conflict or combat—most prominently “fight” (and close synonyms like “brawl”).

    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
    </h2>
    -2.89
     –
    -2.69
    <em>
    -2.67
    とはいえ
    -2.58
    </sup>
    -2.42
        
    -2.39
    趼
    -2.38
     unnecessarily
    -2.31
    鸲
    -2.31
     \
    -2.25
    POSITIVE LOGITS
     kemer
    2.81
    腠
    2.69
     sabun
    2.67
     reigning
    2.56
     munc
    2.55
     uſed
    2.55
     sostu
    2.53
     corriger
    2.52
     diki
    2.48
     blanda
    2.48
    Activations Density 0.005%

    No Known Activations