© 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. 34-GEMMASCOPE-RES-131K
    7. 106197
    Prev
    Next
    INDEX
    Explanations

    assembly registers and specific phrases

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

    The neuron activates on numeric tokens (digits and numbers) in code, i.e. numeric literals and addresses.

    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_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     even
    -1.25
     especially
    -1.17
     their
    -1.17
     both
    -1.16
     particularly
    -1.10
     an
    -0.99
     while
    -0.94
     this
    -0.90
     what
    -0.85
     or
    -0.84
    POSITIVE LOGITS
     Ogni
    0.98
    があるので
    0.97
    咂
    0.95
     ソファ
    0.94
     quelques
    0.92
     maniere
    0.92
    KAN
    0.91
     Pizarro
    0.90
    希望能
    0.90
     Bruxelles
    0.89
    Activations Density 0.015%

    No Known Activations