© 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. 10-GEMMASCOPE-RES-131K
    4. 130760
    Prev
    Next
    INDEX
    Explanations

    chemical and biological processes

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

    scientific or chemical terms with specific suffixes like "emia," "ized," "ampled," or "orylation."

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron responds to long, technical words by spotting their morphemic parts—especially common prefixes or suffixes (e.g. re-parameter-ize, under-sample-ed, hyper-insulinem-ic, sub-stitut-ed, phosph-orylation).

    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
     possíveis
    -1.41
    ‘‘
    -1.33
    certainly
    -1.27
     sustituir
    -1.23
    necessarily
    -1.20
     semelhantes
    -1.19
    縢
    -1.19
    obtained
    -1.19
     praktische
    -1.18
     kuah
    -1.18
    POSITIVE LOGITS
    SceneManagement
    1.29
     up
    1.27
     själv
    1.21
     penyimpanan
    1.20
     بعض
    1.19
     out
    1.16
     setColor
    1.16
     seu
    1.14
     Some
    1.13
     hanya
    1.13
    Activations Density 0.036%

    No Known Activations