© 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. 24-GEMMASCOPE-2-RES-16K
    7. 5117
    Prev
    Next
    INDEX
    Explanations

    equations for length

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

    The neuron fires most strongly on discourse‐structuring tokens—especially ordinals and quantifiers (one, another, third) and gerund or participial verbs (using, referred) that introduce or organize points.

    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
    ը
    0.79
    மிய
    0.77
    katkan
    0.77
     ingres
    0.76
    asyonu
    0.74
    ivamente
    0.73
     ettha
    0.73
    ární
    0.73
    ServiceImpl
    0.72
    徇
    0.72
    POSITIVE LOGITS
    ौनक
    0.84
    𝘽
    0.79
    ກັບ
    0.76
    現代
    0.75
    𝗚
    0.74
    ঘাত
    0.73
    𝙎
    0.70
    𝗙
    0.70
     équip
    0.70
    𝓛
    0.69
    Activations Density 0.000%

    No Known Activations