© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 21327
    Prev
    Next
    INDEX
    Explanations

    news, media, and names

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

    The neuron is picking out numeric metadata in article headers—especially dates, times, and other standalone numbers.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     required
    -1.41
     require
    -1.28
     would
    -1.27
    他能
    -1.22
     key
    -1.21
    此前
    -1.20
     remain
    -1.20
     p
    -1.20
    他要
    -1.17
    自己是
    -1.16
    POSITIVE LOGITS
     mereka
    1.45
     薬局
    1.41
     sekaligus
    1.28
     demokra
    1.27
     actrices
    1.27
    fjspx
    1.23
     gånger
    1.22
    ֺ
    1.21
     them
    1.20
     kutu
    1.20
    Activations Density 0.186%

    No Known Activations