© 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. 11850
    Prev
    Next
    INDEX
    Explanations

    class definition starts

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

    This neuron responds to programming‐style tokens—identifiers, keywords, numbers, and other code syntax elements.

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

    technical documentation and code with phrases about data types, structures, and implementation details.

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

    descriptions and references to programming APIs and data structures, especially code-like type/method declarations and library/interface documentation.

    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
     ESPN
    0.61
     Providing
    0.56
    ularity
    0.56
     डिज
    0.55
     webcast
    0.55
    üb
    0.54
    ód
    0.54
     gin
    0.54
    ESPN
    0.54
     notar
    0.53
    POSITIVE LOGITS
    KeyValuePair
    0.86
     TreeNode
    0.84
    Person
    0.81
     Person
    0.80
    TreeNode
    0.80
     Noeud
    0.74
    Nodo
    0.73
    Node
    0.71
     Node
    0.71
     MenuItem
    0.69
    Activations Density 0.027%

    No Known Activations