© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 84033
    Prev
    Next
    INDEX
    Explanations

    months and punctuation

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

    The neuron activates on numeric tokens and numeric expressions (dates, times, measurements, counts, etc.).

    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
     this
    -1.67
     that
    -1.43
     demselben
    -1.41
     as
    -1.33
     jenem
    -1.23
     and
    -1.20
     these
    -1.16
     hetzelfde
    -1.15
     have
    -1.13
     those
    -1.13
    POSITIVE LOGITS
     murale
    1.68
     solides
    1.55
    recently
    1.55
     élas
    1.54
     principes
    1.51
     alho
    1.48
     tristes
    1.45
     manier
    1.45
     tropicales
    1.44
     jaunes
    1.43
    Activations Density 0.234%

    No Known Activations