© 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. 5766
    Prev
    Next
    INDEX
    Explanations

    in Python code

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

    This neuron fires on salient content words—especially nouns, titles, and key modifiers—that carry the main topic or informative load of a sentence.

    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
     devenir
    0.71
    ︱
    0.64
     sommige
    0.62
     российского
    0.61
     ட
    0.60
     הא
    0.59
    可以说是
    0.59
     diventare
    0.58
    ϊ
    0.57
     সূত্রে
    0.57
    POSITIVE LOGITS
     ফরজ
    0.61
     bhikkhu
    0.58
    CodeableConcept
    0.54
    オブジェクト
    0.54
     elytra
    0.53
     motivic
    0.53
    innerHTML
    0.52
     rune
    0.51
    FirstOrDefault
    0.51
     stator
    0.50
    Activations Density 10.785%

    No Known Activations