© 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. 22-GEMMASCOPE-RES-131K
    7. 71511
    Prev
    Next
    INDEX
    Explanations

    should what would

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

    The neuron is marking tokens that belong to the main body of an article (i.e. real content words) rather than metadata, headers, or other boilerplate.

    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
     dentelle
    -1.47
     esclavos
    -1.45
     renouvel
    -1.39
     einander
    -1.38
     fyra
    -1.35
    gehende
    -1.34
    睃
    -1.25
     nästan
    -1.24
     meistens
    -1.24
     peluche
    -1.22
    POSITIVE LOGITS
     should
    2.14
     what
    1.59
     would
    1.36
    应该
    1.29
     could
    1.28
     must
    1.27
    应该是
    1.27
     seems
    1.25
     seem
    1.24
    似乎
    1.23
    Activations Density 0.022%

    No Known Activations