© 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. 10-GEMMASCOPE-RES-131K
    7. 40692
    Prev
    Next
    INDEX
    Explanations

    arguments and fights

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

    The neuron detects words referring to conflicts or disputes (e.g., argument, fight, quarrel).

    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
     to
    -1.89
     –
    -1.68
        
    -1.62
    नलोड
    -1.62
                
    -1.59
     appear
    -1.52
                      
    -1.52
     ''
    -1.50
               
    -1.48
     has
    -1.47
    POSITIVE LOGITS
     about
    1.79
     zwarte
    1.55
     tuvimos
    1.51
     geschlagen
    1.50
     bezw
    1.48
     bygget
    1.45
    ꞌ
    1.45
     troisième
    1.43
     verdaderas
    1.39
    ܔ
    1.38
    Activations Density 0.016%

    No Known Activations