© 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. 34-GEMMASCOPE-RES-131K
    4. 116918
    Prev
    Next
    INDEX
    Explanations

    usage of the word wet

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

    The neuron consistently activates on the sequence “wet” (either standing alone or as the start of compound forms like “wetting,” “wet‐lab,” “wet‐multi,” etc.).

    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
    modif
    -0.84
     deutschland
    -0.82
    classy
    -0.81
    crocs
    -0.81
    Sentiment
    -0.79
     thấm
    -0.77
     acceptor
    -0.76
    skyline
    -0.76
     chrys
    -0.75
    closeup
    -0.75
    POSITIVE LOGITS
     Wet
    1.77
    Wet
    1.70
     wet
    1.59
    wet
    1.56
    zel
    1.13
    tability
    1.08
    suit
    1.03
    ump
    1.00
     WET
    0.98
    ted
    0.93
    Activations Density 0.009%

    No Known Activations