© 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. 40190
    Prev
    Next
    INDEX
    Explanations

    rounding to nearest

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

    The neuron activates on numerical tokens (whole numbers or multi‐digit numbers) in mathematical questions.

    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
    You
    -1.29
     only
    -1.27
     岁
    -1.21
    The
    -1.20
    only
    -1.16
    People
    -1.14
    ย์
    -1.13
     new
    -1.13
    even
    -1.11
     vib
    -1.11
    POSITIVE LOGITS
    Cie
    1.26
     damal
    1.25
     tatsächlich
    1.23
    Struktur
    1.22
    odziel
    1.20
    ᄏ
    1.20
    んですね
    1.19
     meestal
    1.18
    割と
    1.16
     refroid
    1.15
    Activations Density 0.042%

    No Known Activations