© 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. 1861
    Prev
    Next
    INDEX
    Explanations

    approximate numbers

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

    The neuron is strongly activated by numeric tokens (quantities, dates, IDs, file‐path numbers), i.e. it detects numbers.

    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
     дані
    0.34
     Faktoren
    0.30
    Authentication
    0.29
    Appending
    0.29
    Replace
    0.28
     maxLength
    0.28
    ゙
    0.28
    っき
    0.27
    Append
    0.27
    Przypisy
    0.27
    POSITIVE LOGITS
    ;
    0.38
     Carolina
    0.37
    trombone
    0.36
     fledgling
    0.36
    ri
    0.34
    spired
    0.33
     lemonade
    0.33
    .,
    0.33
    .;
    0.32
    ashington
    0.32
    Activations Density 2.254%

    No Known Activations