© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 128539
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on tokens containing numeric values (e.g. years, enzyme codes, measurements, or other digit-based identifiers).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    alternative names or specifications

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     and
    -1.19
     which
    -1.04
     between
    -1.01
     kdo
    -1.00
    antism
    -0.88
    >);
    -0.85
    ᧐
    -0.85
     који
    -0.84
    henswürdigkeiten
    -0.82
     lacked
    -0.81
    POSITIVE LOGITS
    AKA
    1.08
    字典
    0.98
    aka
    0.96
     bénéfice
    0.96
     diamants
    0.96
     Produit
    0.91
     incendie
    0.88
     amélior
    0.87
    监听
    0.86
    各种
    0.85
    Activations Density 0.126%

    No Known Activations