© 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
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 82942
    Prev
    Next
    INDEX
    Explanations

    verbs with adverbs or prepositions

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

    The neuron detects words in idiomatic expressions that convey cognitive or emotional impact—verbs and phrases like “ring true,” “hit home,” “weigh heavily,” or “ring a bell.”

    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
    '">
    -0.95
    があります
    -0.93
    relu
    -0.89
     Θ
    -0.86
    這
    -0.83
    出现的
    -0.82
    billon
    -0.81
    ameter
    -0.80
    ('|
    -0.79
    ९
    -0.79
    POSITIVE LOGITS
     strong
    1.27
     strongly
    1.25
     itself
    1.25
     heavily
    1.16
     differently
    1.12
    Throughout
    1.02
     heavy
    1.00
     well
    0.98
     prominently
    0.97
     loudly
    0.97
    Activations Density 0.036%

    No Known Activations