© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 72600
    Prev
    Next
    INDEX
    Explanations

    urls and ampersands

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

    The neuron strongly activates on long runs of digits (often mixed with punctuation), i.e. densely packed numeric data sequences.

    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
    сезон
    -1.24
     في
    -1.16
     atau
    -1.16
     یا
    -1.15
    Although
    -1.14
    During
    -1.13
     Aside
    -1.10
     Although
    -1.10
     принад
    -1.09
     Regardless
    -1.09
    POSITIVE LOGITS
     kinda
    1.13
     함께
    1.09
     configuración
    1.07
     !!!!
    1.06
    !!!!!!!
    1.06
     sooo
    1.04
     работают
    1.03
     имеются
    1.03
     mauva
    1.02
     ludicrous
    1.02
    Activations Density 0.030%

    No Known Activations