© 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. 71753
    Prev
    Next
    INDEX
    Explanations

    references to drawings or figures

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

    The neuron is keyed to spotting numeric tokens (i.e. any sequence of digits such as section numbers, figure numbers, years, or other standalone numbers).

    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
     perfekte
    -1.15
    |$.
    -1.12
     embarazadas
    -1.11
     různých
    -1.08
     &
    -1.05
     hermosa
    -1.02
     similar
    -1.02
     your
    -1.01
     nagyon
    -1.00
    TextWatcher
    -1.00
    POSITIVE LOGITS
     volant
    1.16
     tass
    1.13
    曲を
    1.12
    ⁉
    1.12
     perle
    1.11
     enne
    1.10
     anta
    1.10
     vana
    1.09
     gammal
    1.06
     amat
    1.05
    Activations Density 0.173%

    No Known Activations