© 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. 10-GEMMASCOPE-RES-131K
    7. 103908
    Prev
    Next
    INDEX
    Explanations

    ending code blocks

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

    closing braces or brackets at the end of code blocks, particularly in JavaScript functions and control structures.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @bfgcn7frf5

    The neuron is detecting numeric literals (i.e. numbers and number-like tokens) in the code.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     about
    -1.64
     such
    -1.46
     outras
    -1.44
    蝼
    -1.43
     sujeito
    -1.43
     hierzu
    -1.40
     concludes
    -1.38
     otras
    -1.37
     ultimately
    -1.37
     recientemente
    -1.36
    POSITIVE LOGITS
    (
    1.72
    All
    1.59
    ,
    1.53
    The
    1.52
    Any
    1.48
    +
    1.43
    Being
    1.41
    If
    1.41
    他们
    1.36
    K
    1.35
    Activations Density 0.005%

    No Known Activations