© 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. 36247
    Prev
    Next
    INDEX
    Explanations

    Iraq in, procedure if, Posts tagged

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

    The neuron fires on standalone numeric values and all-caps tokens (e.g. version numbers, acronyms, constants).

    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
     in
    -1.05
    年底
    -1.01
    umum
    -1.00
     their
    -0.98
     rapidly
    -0.96
     or
    -0.94
     make
    -0.93
     quickly
    -0.92
     that
    -0.92
     should
    -0.88
    POSITIVE LOGITS
    粜
    1.05
     имел
    1.03
    翃
    1.02
     raconter
    1.00
     thậm
    0.96
     enlight
    0.94
     emporter
    0.93
    automne
    0.92
    耷
    0.92
    municipio
    0.91
    Activations Density 0.000%

    No Known Activations