© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 2441
    Prev
    Next
    INDEX
    Explanations

    This neuron detects procedural or instructional language—imperative steps and commands describing programming tasks.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    articles ("the") in technical or instructional writing contexts.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    step-by-step technical instructions in programming tutorials, especially imperative directives describing actions to perform (e.g., creating methods/services, fetching or populating data, and procedural steps).

    oai_token-act-pair · gpt-5Triggered 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
     Defaults
    0.58
     mActivity
    0.55
     Firebase
    0.50
     inactive
    0.49
     defaults
    0.49
     Canvas
    0.48
     supabase
    0.48
     Panel
    0.47
     ios
    0.47
     default
    0.46
    POSITIVE LOGITS
    helloworld
    0.83
    HelloWorld
    0.68
    hello
    0.63
    java
    0.60
    Hello
    0.57
    HELLO
    0.56
    student
    0.52
    JDBC
    0.52
    arrayList
    0.51
    respuesta
    0.49
    Activations Density 0.002%

    No Known Activations