© 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. 2020
    Prev
    Next
    INDEX
    Explanations

    high quality or quantity descriptors

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

    The neuron responds to strong degree‐ or quantity‐expressing words (intensifiers and quantifiers) such as “high,” “lot (of),” “mixed,” “good,” 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
     ayant
    -1.22
     häl
    -1.09
    AndGet
    -1.08
    efforts
    -1.08
     esfuerzo
    -1.05
     primeiros
    -1.04
    effort
    -1.02
    rige
    -1.02
    ära
    -1.01
     effort
    -0.98
    POSITIVE LOGITS
     problems
    1.45
     nodig
    1.38
     experience
    1.38
     tools
    1.34
     access
    1.34
     ideas
    1.25
     concerns
    1.24
     knowledge
    1.23
     ambitions
    1.16
     options
    1.16
    Activations Density 0.114%

    No Known Activations