© 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. 11831
    Prev
    Next
    INDEX
    Explanations

    ```sql` and ````python` fence

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

    This neuron fires on the “+” diff‐addition marker and the shell-command tokens (e.g. “echo”) that follow it, i.e. added lines in a patch/diff.

    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
    ِس
    0.62
    、
    0.55
    Kale
    0.54
    veil
    0.53
    elevationMap
    0.53
    Said
    0.52
    곧
    0.52
    鉉
    0.50
    fileExists
    0.50
    rots
    0.49
    POSITIVE LOGITS
       
    1.48
     ，
    0.93
     ”
    0.81
     ”.
    0.73
     ”,
    0.72
           
    0.70
     ،
    0.69
               
    0.63
    
    0.63
     LayoutInflater
    0.61
    Activations Density 0.036%

    No Known Activations