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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 33183
    Prev
    Next
    INDEX
    Explanations

    medical and scientific suffixes

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

    scientific and medical terms that end with suffixes like "onic", "phic", "inal", "omic", or "agic".

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

    The neuron flags domain‐specific technical adjectives—especially those formed with Greek/Latin suffixes like “-ic” or “-al.”

    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
    ?
    -1.41
    </h4>
    -1.24
    で
    -1.16
    啖
    -1.15
     empêcher
    -1.13
     '^
    -1.11
     &
    -1.10
     klassischen
    -1.09
    l
    -1.08
     быть
    -1.07
    POSITIVE LOGITS
     that
    1.52
     PUSTAKA
    1.42
    useNewUrlParser
    1.41
    本章
    1.37
     achieves
    1.36
     στι
    1.34
    发布于
    1.32
    洢
    1.31
     leveraging
    1.30
    gantung
    1.29
    Activations Density 0.278%

    No Known Activations