© 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. 7810
    Prev
    Next
    INDEX
    Explanations

    Welsh places and names

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

    The neuron is picking out numeric tokens (especially figures, rankings, years, and other quantitative values) in the text.

    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
     Kremlin
    0.55
     handles
    0.54
     Novgorod
    0.54
     Moscow
    0.54
    괵
    0.54
     Albrecht
    0.53
     KX
    0.53
     consciousness
    0.53
     আস
    0.52
    ^{+}$
    0.52
    POSITIVE LOGITS
     Welsh
    1.89
    Wales
    1.88
     Wales
    1.75
     wales
    1.49
     Cardiff
    1.47
     Llan
    1.24
     Swansea
    1.23
     Cym
    1.14
     Snowdon
    1.02
    󠁷
    0.99
    Activations Density 0.111%

    No Known Activations