© 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. 58680
    Prev
    Next
    INDEX
    Explanations

    fighting

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

    The neuron spotlights words denoting physical combat matchups—most notably “Fight” (and variants like “brawl”) in the names or descriptions of fighting events.

    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
    ünster
    -1.02
    matische
    -0.89
     dán
    -0.83
     Expan
    -0.81
    </h2>
    -0.80
    rentiel
    -0.80
    生意
    -0.79
    dzić
    -0.78
     Anis
    -0.77
    过来
    -0.76
    POSITIVE LOGITS
     Fight
    1.79
     fight
    1.50
    fight
    1.49
     Fighting
    1.39
     FIGHT
    1.38
    Fight
    1.33
     fights
    1.25
    FIGHT
    1.24
    Fighting
    1.20
     Club
    1.15
    Activations Density 0.011%

    No Known Activations