© 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. 14329
    Prev
    Next
    INDEX
    Explanations

    unified systems and single solutions

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

    The neuron fires strongly on participial or compound adjectives (often hyphenated or ending in “-ed”) that describe states, qualities, or specialized conditions.

    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
     Ot
    0.60
     (“
    0.60
     multim
    0.57
    and
    0.57
    的主要
    0.56
    önt
    0.56
     (=
    0.55
     Coll
    0.55
    ‒
    0.55
    ‌هایی
    0.53
    POSITIVE LOGITS
     شدن
    0.92
    പ്പെട
    0.77
    打造
    0.76
     ঘোষনা
    0.76
    IZATION
    0.75
    ization
    0.74
    化
    0.73
    izada
    0.72
    лизация
    0.71
    这句话
    0.71
    Activations Density 0.143%

    No Known Activations