© 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. 14332
    Prev
    Next
    INDEX
    Explanations

    mathematical expressions end

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

    The neuron fires specifically on those odd “)\ 248”, “)\ 250”, “)\ 252”, etc. tokens—that is, closing‐parenthesis tokens carrying numeric escape codes in LaTeX math mode.

    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
     bekannte
    -2.73
    }
    -2.63
     is
    -2.61
    L
    -2.53
    ]
    -2.50
     französische
    -2.30
     eigentliche
    -2.22
    骘
    -2.20
     sogenannte
    -2.14
     постепенно
    -2.14
    POSITIVE LOGITS
    </strong>
    3.56
    ᛬
    2.69
    
    2.67
    皤
    2.56
    屣
    2.42
     we
    2.39
    狁
    2.38
    恹
    2.36
    as
    2.34
    
    2.34
    Activations Density 0.002%

    No Known Activations