© 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. 10-GEMMASCOPE-RES-131K
    7. 90812
    Prev
    Next
    INDEX
    Explanations

    concepts after 'and'

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

    The neuron is tuned to spotting numeric quantity tokens (especially decimals and measurement values).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
    various
    -1.52
     utilizzando
    -1.43
    Utiliz
    -1.43
     bebés
    -1.41
    ꯷
    -1.39
    ående
    -1.38
    يست
    -1.37
     meravigli
    -1.36
     verschillende
    -1.35
    iyle
    -1.35
    POSITIVE LOGITS
     then
    1.73
     who
    1.67
     his
    1.59
     at
    1.53
     could
    1.48
     before
    1.48
     this
    1.45
     just
    1.45
     during
    1.38
     has
    1.37
    Activations Density 0.232%

    No Known Activations