© 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. 10783
    Prev
    Next
    INDEX
    Explanations

    aerial or aloft

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

    The neuron detects numeric tokens—especially digits and decimal numbers—i.e. parts of measurements or numeric values.

    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
     cyclospor
    0.67
    潺
    0.55
     jdbcTemplate
    0.55
    吲
    0.55
     பின்ன
    0.54
     transplant
    0.53
     Lorena
    0.53
     Sql
    0.52
     %@",
    0.52
    श्याम
    0.52
    POSITIVE LOGITS
     aloft
    1.66
     aerial
    1.50
    空中
    1.48
     atop
    1.45
     rooftop
    1.43
     Aerial
    1.41
    aerial
    1.41
     tower
    1.34
     heights
    1.34
    Aerial
    1.33
    Activations Density 0.684%

    No Known Activations