© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 13193
    Prev
    Next
    INDEX
    Explanations

    transition from X to Y

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

    This neuron is highly activated by the preposition “from” (and similarly “to”), essentially detecting the “from … to” pattern that marks 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_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
    瞟
    -1.48
    皤
    -1.41
     nimmer
    -1.38
     dahil
    -1.34
    粜
    -1.31
     passende
    -1.30
    рочка
    -1.30
     ↓,
    -1.30
     siè
    -1.28
     พ
    -1.27
    POSITIVE LOGITS
     that
    1.61
     During
    1.55
     at
    1.54
     this
    1.48
     just
    1.45
     what
    1.36
     probably
    1.33
     not
    1.32
     only
    1.32
     where
    1.29
    Activations Density 0.068%

    No Known Activations