© 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. 10-GEMMASCOPE-RES-131K
    4. 72907
    Prev
    Next
    INDEX
    Explanations

    absolute value and integrals

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

    The neuron strongly activates on numerals (numeric constants or magnitudes) in mathematical expressions, particularly when they appear inside LaTeX‐style math delimiters.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     :]
    -2.05
     内衣
    -2.02
     .=
    -2.02
    atigable
    -1.98
    儇
    -1.95
     гель
    -1.88
    ات
    -1.84
     ۱
    -1.82
     ≤
    -1.82
    b
    -1.81
    POSITIVE LOGITS
    ↵
    2.86
     and
    2.39
     they
    2.28
     which
    1.73
    germany
    1.63
     inicialmente
    1.61
    前方
    1.57
     of
    1.57
     continúa
    1.56
     शुरु
    1.55
    Activations Density 0.005%

    No Known Activations