© 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
    3. Gemma-3-12B
    5. 31-GEMMASCOPE-2-RES-16K
    7. 15793
    Prev
    Next
    INDEX
    Explanations

    The neuron fires on numeric tokens, especially numbers with digits (including decimal points).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    phrases indicating that something is subject to change or modification.

    oai_token-act-pair · claude-4-5-sonnetTriggered by @jyhe0408

    language indicating updates, edits, revisions, or changes over time, especially in formal policy or terms contexts.

    oai_token-act-pair · gpt-5Triggered 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
    Reduction
    0.59
     снижение
    0.56
     eliminación
    0.54
    Reduce
    0.54
     elimination
    0.54
     Reduction
    0.52
    stieg
    0.52
    を残
    0.51
     reduction
    0.50
     लेने
    0.50
    POSITIVE LOGITS
     changed
    3.14
     changing
    3.02
    changed
    2.92
     changes
    2.91
     Changed
    2.91
     변경
    2.89
     تغییر
    2.81
     change
    2.78
    changing
    2.78
    変更
    2.78
    Activations Density 0.838%

    No Known Activations