© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 32650
    Prev
    Next
    INDEX
    Explanations

    numbers following H

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

    The neuron is highly activated by numeral tokens (digits or multi-digit numbers, e.g. years and other numeric sequences).

    oai_token-act-pair · o4-mini
    Triggered 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
    Negative Logits
     to
    -1.21
     unusual
    -0.98
    合った
    -0.96
     any
    -0.95
     result
    -0.93
     without
    -0.92
    MatInputModule
    -0.91
    MatIconModule
    -0.91
     failed
    -0.89
     aimed
    -0.89
    POSITIVE LOGITS
     kapture
    1.13
     fote
    1.03
     meubles
    0.99
    peen
    0.99
     nader
    0.96
     gewi
    0.94
     CAI
    0.94
     yves
    0.93
     encou
    0.92
     Perché
    0.91
    Activations Density 0.001%

    No Known Activations