© 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. 64
    Prev
    Next
    INDEX
    Explanations

    for the

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

    The neuron fires on words describing configuration parameters or default settings—especially in “used for” or “default” clauses—in technical documentation.

    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
     escova
    -1.07
    )}]{
    -1.05
     тогда
    -1.05
     requisite
    -1.05
     vijf
    -1.04
     banho
    -1.02
    ブログ村
    -1.02
     dangers
    -1.01
     Jego
    -1.00
     puramente
    -1.00
    POSITIVE LOGITS
     when
    1.81
     during
    1.61
     your
    1.59
     each
    1.41
     will
    1.38
     if
    1.36
     throughout
    1.28
    Each
    1.24
     nantinya
    1.23
    when
    1.20
    Activations Density 0.044%

    No Known Activations