© 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. 34-GEMMASCOPE-RES-131K
    7. 83986
    Prev
    Next
    INDEX
    Explanations

    citation markers like `[^1]`

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

    The neuron selectively activates on standalone numeric tokens—especially reference or footnote numbers and citation indices—marking places where numbers serve as superscripted or bracketed citations.

    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
     České
    -0.94
    ],"
    -0.90
    liminal
    -0.88
    thday
    -0.86
    doll
    -0.85
    ادم
    -0.84
    ;;)
    -0.84
    mphs
    -0.81
     various
    -0.81
    ğından
    -0.80
    POSITIVE LOGITS
    </sup>
    1.27
    </strong>
    1.09
    ued
    0.97
    </em>
    0.97
     difficiles
    0.89
    herent
    0.88
    oras
    0.87
    Close
    0.86
    </blockquote>
    0.84
     coû
    0.84
    Activations Density 0.004%

    No Known Activations