© 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. 34-GEMMASCOPE-RES-131K
    4. 86895
    Prev
    Next
    INDEX
    Explanations

    appropriate followed by context

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

    The neuron fires on the word “appropriate” (and its variants like “inappropriate”) whenever it marks suitability or correctness in the text.

    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
    ’,
    -1.46
     surpass
    -1.21
    работ
    -1.14
    étaires
    -1.13
    戋
    -1.11
     hacen
    -1.10
    怎么样
    -1.09
    ování
    -1.09
    굽
    -1.08
    rophilic
    -1.07
    POSITIVE LOGITS
     sized
    1.39
     amounts
    1.12
     levels
    1.07
    According
    1.06
    eapp
    1.05
    垍
    1.05
    となりました
    1.04
     level
    1.02
    U
    1.00
     for
    0.99
    Activations Density 0.017%

    No Known Activations