© 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. 22-GEMMASCOPE-RES-131K
    7. 14118
    Prev
    Next
    INDEX
    Explanations

    New followed by a

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

    This neuron detects standalone document-structure numbers—particularly section or chapter numerals (e.g. Roman-numeral headings like “V” or “C”) and other isolated numeric tokens used for numbering.

    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
     histoires
    -1.35
     \%}$
    -1.20
     baratas
    -1.18
     Transporte
    -1.18
     Alimentos
    -1.14
     Português
    -1.13
     Europas
    -1.11
     Arquite
    -1.10
     آیفون
    -1.09
     Sábado
    -1.09
    POSITIVE LOGITS
    =[];
    1.23
     erziel
    1.16
     acogedor
    1.09
    []){
    1.09
     erhielt
    1.09
     телефона
    1.07
     erreichte
    1.06
     tumbuhan
    1.03
    Gön
    1.03
    шений
    1.02
    Activations Density 0.001%

    No Known Activations