© 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. 8278
    Prev
    Next
    INDEX
    Explanations

    Rust code snippets

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

    The neuron strongly activates on four‐digit numbers, especially publication years in reference citations.

    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
     mut
    0.58
    mut
    0.55
     ज्‍यादा
    0.54
     Mut
    0.53
     অ্যাল
    0.53
    篇
    0.50
    (`#
    0.49
    Mut
    0.49
    պես
    0.49
    VN
    0.49
    POSITIVE LOGITS
     Ok
    0.66
    Ok
    0.65
     ok
    0.64
    ::<
    0.63
     beliebt
    0.58
     alright
    0.55
     facile
    0.54
     clinch
    0.54
     okay
    0.53
     negro
    0.53
    Activations Density 0.005%

    No Known Activations