© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-12B
    5. 24-GEMMASCOPE-2-RES-16K
    7. 955
    Prev
    Next
    INDEX
    Explanations

    The neuron fires strongly on four‐digit numbers with decimal fractions (e.g. “2004.8728,” “1724.4202,” etc.), i.e. numeric citation or publication‐year 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
    ,
    0.44
    ق
    0.41
    ной
    0.40
    TING
    0.40
    ר
    0.39
    ो
    0.38
    п
    0.37
    í
    0.37
    ح
    0.37
    ──
    0.36
    POSITIVE LOGITS
     illetve
    0.61
     czyli
    0.52
     odnosno
    0.52
     valamint
    0.51
     który
    0.51
     jonka
    0.50
     která
    0.49
     který
    0.48
     cujo
    0.48
     ktoré
    0.47
    Activations Density 0.000%

    No Known Activations