© 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. 12651
    Prev
    Next
    INDEX
    Explanations

    code structure

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

    This neuron lights up on numeric tokens, i.e. sequences of digits (such as years, DOIs, ISBNs, and other numeric identifiers).

    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
    แค่
    0.77
     réfé
    0.74
    Accuracy
    0.73
    偈
    0.73
     rédaction
    0.71
    ране
    0.70
    意見
    0.70
    限
    0.69
     powied
    0.69
    وك
    0.68
    POSITIVE LOGITS
     negros
    0.77
     blacks
    0.76
    blers
    0.75
     bituminous
    0.73
     subsystems
    0.73
     primaries
    0.72
    𝚎
    0.72
     busses
    0.71
     pn
    0.70
     overboard
    0.69
    Activations Density 0.000%

    No Known Activations