© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 26123
    Prev
    Next
    INDEX
    Explanations

    sequence following punctuation

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

    The neuron responds to sentence boundaries, peaking on end-of-sentence punctuation (e.g. “.”) and the first tokens of the next sentence.

    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
    -0.91
    lemme
    -0.90
     from
    -0.89
     Sha
    -0.84
     früher
    -0.84
     if
    -0.84
     will
    -0.83
     doesn
    -0.81
    են
    -0.79
     helps
    -0.79
    POSITIVE LOGITS
     followed
    1.41
     then
    1.38
     Subsequently
    1.30
    那时
    1.26
     subsequently
    1.26
     thereafter
    1.26
    Later
    1.22
     later
    1.22
     subsequent
    1.21
    同年
    1.17
    Activations Density 0.101%
    No Known Activations