© 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. 9388
    Prev
    Next
    INDEX
    Explanations

    numbers followed by punctuation

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

    This neuron consistently activates on numeric tokens—particularly version numbers, line‐count markers, and other numerical identifiers found in patch/diff headers and similar metadata.

    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
     that
    -1.35
     should
    -1.16
     for
    -1.13
     Before
    -1.07
     to
    -1.02
     However
    -0.98
     from
    -0.96
     its
    -0.96
    [{{
    -0.96
     in
    -0.95
    POSITIVE LOGITS
    łk
    1.14
     nœ
    1.02
     élas
    1.02
     levier
    1.00
     résine
    0.98
     murale
    0.97
     étan
    0.97
     cocinas
    0.93
    ቸ
    0.91
     provoquer
    0.90
    Activations Density 0.000%

    No Known Activations