© 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. 22-GEMMASCOPE-RES-131K
    7. 12837
    Prev
    Next
    INDEX
    Explanations

    supported by, also funded

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

    The neuron is picking up on formatted reference/citation markers—periods, commas, and numeric tokens typically found in bibliographies and acknowledgments.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     Pues
    -1.25
    Pues
    -1.08
    Especificações
    -1.07
    닌
    -1.04
     Entonces
    -1.00
    Conclusão
    -1.00
    hears
    -1.00
     upang
    -1.00
    anmar
    -0.98
    watched
    -0.98
    POSITIVE LOGITS
     and
    1.46
     также
    1.20
     también
    1.10
    
    1.02
    Geld
    0.97
     myös
    0.97
    Tja
    0.97
    Doktor
    0.96
    همچنین
    0.96
     additionally
    0.95
    Activations Density 0.039%

    No Known Activations