© 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. 54276
    Prev
    Next
    INDEX
    Explanations

    various prefixes and suffixes

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

    The neuron strongly activates on digit tokens and numeric sequences, i.e. it flags occurrences of 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
     from
    -1.94
     only
    -1.68
     several
    -1.52
     every
    -1.47
     by
    -1.46
     because
    -1.34
     one
    -1.30
     tagit
    -1.28
     like
    -1.28
     had
    -1.27
    POSITIVE LOGITS
    仍然
    1.34
     verranno
    1.30
     potranno
    1.28
    *;
    1.27
    seven
    1.25
    −/−
    1.24
    ___________
    1.21
     ____________
    1.18
    — 
    1.18
    ources
    1.18
    Activations Density 0.010%

    No Known Activations