© 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. 117428
    Prev
    Next
    INDEX
    Explanations

    numbers and identifiers

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

    The neuron strongly activates on tokens that are (multi‐digit) numeric strings—e.g. years, citation numbers, code/ID numbers.

    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
     dozens
    -0.81
     She
    -0.76
    xffffffff
    -0.75
    xFFFFFFFF
    -0.70
    anghai
    -0.69
    afel
    -0.66
    netti
    -0.66
     While
    -0.65
    回復
    -0.65
    شع
    -0.65
    POSITIVE LOGITS
    thousands
    1.30
     thousands
    1.18
     milliers
    0.89
     tens
    0.89
     THOUS
    0.88
    三千
    0.88
     lakhs
    0.87
    XXXXX
    0.86
    Thousands
    0.86
     nich
    0.80
    Activations Density 0.428%

    No Known Activations