© 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. 98599
    Prev
    Next
    INDEX
    Explanations

    verbs and their common endings

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

    The neuron fires on long, Latinate verb stems or formal scientific/action words (e.g. “adjour-,” “prioriti-,” “quantit-,” “super-,” “scrutin-,” “impair-,” “detem-,” “synchron-,” “destabili-”).

    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
    などは
    -1.25
     and
    -1.22
    そんな
    -1.19
     lã
    -1.17
     however
    -1.17
    なども
    -1.16
    もありますが
    -1.13
     &
    -1.13
    прочем
    -1.12
     però
    -1.09
    POSITIVE LOGITS
     it
    2.16
     became
    1.95
     was
    1.58
     gave
    1.50
     will
    1.50
     took
    1.49
     began
    1.48
     went
    1.38
     showed
    1.38
     provides
    1.35
    Activations Density 0.025%

    No Known Activations