© 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. 34-GEMMASCOPE-RES-131K
    7. 63323
    Prev
    Next
    INDEX
    Explanations

    years and ranges

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

    The neuron fires on numeric tokens and numeric sequences—especially years, decades, and number ranges.

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     when
    -1.41
     with
    -1.24
     make
    -1.22
     because
    -1.20
     bringing
    -1.08
     while
    -1.06
     will
    -1.05
     if
    -1.04
     after
    -1.03
     those
    -1.01
    POSITIVE LOGITS
     alluring
    1.35
    pendentif
    1.27
    GOTREF
    1.19
     sourire
    1.17
     récemment
    1.16
    鸝
    1.15
     direc
    1.15
     Atención
    1.13
    isies
    1.13
     Sünde
    1.12
    Activations Density 0.002%

    No Known Activations