© 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. 34-GEMMASCOPE-RES-131K
    7. 84487
    Prev
    Next
    INDEX
    Explanations

    end of expressions or code blocks

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

    The neuron selectively fires on code punctuation that separates statements or elements—primarily commas and semicolons.

    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
    TintMode
    -1.20
     amerik
    -1.20
    ༵
    -1.19
     dáms
    -1.14
     női
    -1.13
     kuv
    -1.12
    敺
    -1.11
    雠
    -1.09
     luon
    -1.09
    nextProps
    -1.08
    POSITIVE LOGITS
     this
    1.13
     every
    1.09
     anyone
    1.04
     any
    1.02
     there
    1.00
     (
    0.94
    chene
    0.94
     instances
    0.91
    this
    0.89
     neither
    0.87
    Activations Density 0.002%

    No Known Activations