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

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 68905
    Prev
    Next
    INDEX
    Explanations

    numbers and units

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

    The neuron is triggered by numeric tokens at the start of multi‐digit figures (e.g. the "3" in “300” days and the "9" in “90‐90‐90” goals).

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

    No Comments

    Negative Logits
    今月
    -1.11
    今週
    -1.09
    つい
    -1.05
    来週
    -1.05
     {
    -1.05
    voie
    -1.03
    もうすぐ
    -1.03
    そろそろ
    -1.02
    --}}
    -1.02
    楽天
    -1.02
    POSITIVE LOGITS
     formales
    1.16
    Derp
    1.15
    fordshire
    1.09
     aparece
    1.08
    tiness
    1.08
     if
    1.07
     tanpa
    1.05
    bodied
    1.04
    しに
    1.04
    リーナ
    1.03
    Activations Density 0.150%

    No Known Activations