© 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. 130750
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on end‐of‐sentence punctuation (periods, question marks, ellipses, etc.), marking sentence boundaries.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    end of sentences

    np_acts-logits-general · gemini-2.5-flash-lite
    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
     this
    -1.50
     Moreover
    -1.50
     Furthermore
    -1.49
    これらの
    -1.48
     Additionally
    -1.41
     بسیاری
    -1.35
     данной
    -1.34
     länder
    -1.31
     tersebut
    -1.30
     Therefore
    -1.27
    POSITIVE LOGITS
     just
    1.72
     doesn
    1.66
     I
    1.66
     they
    1.62
     didn
    1.45
     maybe
    1.44
     kinda
    1.41
     aren
    1.39
     don
    1.35
     víc
    1.34
    Activations Density 0.062%

    No Known Activations