© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW
    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    2. Gemma-3-12B
    3. 24-GEMMASCOPE-2-RES-16K
    4. 14215
    Prev
    Next
    INDEX
    Explanations

    n-grams and mathematical notation

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

    The neuron flags unusually salient or “out‐of‐the‐ordinary” tokens—things like exclamation points, numbers, and other rare or special‐case words.

    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_24_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
    t
    0.88
    من
    0.80
    p
    0.78
    מ
    0.77
    me
    0.76
    på
    0.75
    APP
    0.74
    d
    0.72
    大
    0.72
    ол
    0.71
    POSITIVE LOGITS
    DummyView
    0.85
    speople
    0.80
     subordinates
    0.80
     instructs
    0.79
     conversely
    0.77
     фаразы
    0.77
     endeavours
    0.75
     தெரிவிக்க
    0.73
     grievous
    0.72
    𝙡
    0.72
    Activations Density 0.000%

    No Known Activations