© 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. 98465
    Prev
    Next
    INDEX
    Explanations

    citation numbers, references [@B], [@ref]

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

    The neuron is strongly activated by numerical tokens—digits, measurements, percentages, and reference numbers—in the text.

    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
    是你
    -1.52
     ejecutar
    -1.50
     REALLY
    -1.42
     insanely
    -1.41
     wouldn
    -1.36
    莛
    -1.34
     понадобится
    -1.32
     you
    -1.32
    assies
    -1.29
     ſer
    -1.28
    POSITIVE LOGITS
     many
    1.86
     various
    1.84
     both
    1.80
     mainly
    1.55
     numerous
    1.53
     first
    1.41
     several
    1.38
     for
    1.38
     some
    1.33
     highly
    1.32
    Activations Density 0.127%

    No Known Activations