© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 7137
    Prev
    Next
    INDEX
    Explanations

    The neuron strongly activates on words that appear in diff or patch metadata and comment headings (for example, lines beginning with “+” or “#”), effectively marking tokens in version‐control diffs and comment lines.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-12b-pt/resid_post/layer_31_width_16k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    <strong>
    1.11
    <b>
    0.93
    <code>
    0.92
    .”
    0.79
    <sup>
    0.75
    <em>
    0.74
    <i>
    0.73
     .”
    0.72
     ”.
    0.71
    ,”
    0.71
    POSITIVE LOGITS
     usamos
    1.16
     necesitamos
    1.05
     FIXME
    1.05
     queremos
    1.01
     đc
    0.99
     TODO
    0.98
    TODO
    0.98
     داریم
    0.97
     aqui
    0.96
     kvůli
    0.94
    Activations Density 0.035%

    No Known Activations