© 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
    2. Gemma-2-27B
    3. 10-GEMMASCOPE-RES-131K
    4. 11572
    Prev
    Next
    INDEX
    Explanations

    digits following figure references

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

    The neuron selectively activates on standalone numeric tokens (e.g. figure numbers, reference indices, measurement values).

    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
    :[],
    -1.69
     who
    -1.67
     mengucap
    -1.66
     which
    -1.65
    lmao
    -1.58
     wußte
    -1.57
    :{},
    -1.54
     some
    -1.54
     которые
    -1.53
     when
    -1.52
    POSITIVE LOGITS
     is
    2.64
     are
    2.23
     will
    2.20
     would
    1.63
     must
    1.59
    Your
    1.55
    SaveChangesAsync
    1.46
     рекомендую
    1.43
    是很
    1.41
     
    1.41
    Activations Density 0.254%

    No Known Activations