© 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. 31-GEMMASCOPE-2-RES-16K
    4. 8098
    Prev
    Next
    INDEX
    Explanations

    top up, top topics, top lists, top-tier

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

    The neuron activates on list‐style “Top N” ranking headings (e.g. “Top 5,” “Top 10,” etc.).

    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
     ABOUT
    0.58
    érg
    0.57
     ARISING
    0.56
     about
    0.55
     assuming
    0.55
     presume
    0.54
    仮
    0.54
    捷
    0.52
     shouldn
    0.52
    ủy
    0.52
    POSITIVE LOGITS
    notch
    1.37
     top
    1.16
    স্থানীয়
    1.15
     notch
    1.14
    TOP
    1.10
     Top
    1.09
    Top
    1.08
    top
    1.05
     топо
    1.03
    ographical
    1.02
    Activations Density 0.172%

    No Known Activations