© 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. 31-GEMMASCOPE-2-RES-16K
    4. 317
    Prev
    Next
    INDEX
    Explanations

    The neuron activates on numeric tokens—especially multi‐digit numbers, dates, measurements, and other digit sequences—in 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
    !(
    0.35
    人と
    0.32
    highway
    0.31
    three
    0.31
    social
    0.31
    ...";
    0.31
    what
    0.31
    decorative
    0.30
    社の
    0.30
    historical
    0.30
    POSITIVE LOGITS
    )
    1.41
    ),
    1.38
    )
    1.14
    )،
    1.10
    )/
    1.09
    )\
    1.09
    )*
    1.09
    )+
    1.08
    )...
    1.08
    ),
    1.06
    Activations Density 0.352%

    No Known Activations