© 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. 84195
    Prev
    Next
    INDEX
    Explanations

    code snippets and structured data formats

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

    The neuron activates on numeric literals (especially floating‐point numbers and other digit 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
     two
    -1.21
     only
    -1.18
     just
    -1.17
    pergillus
    -1.15
     what
    -1.15
     on
    -1.13
     those
    -1.12
     three
    -1.09
    corbic
    -1.09
     almost
    -1.07
    POSITIVE LOGITS
     методы
    1.07
     détaillés
    1.05
     testar
    1.05
    ांक
    1.05
     erforder
    1.03
    zucht
    1.02
    eſt
    1.00
     ayudará
    1.00
    Ingrédients
    0.99
    看上去
    0.97
    Activations Density 0.001%

    No Known Activations