© 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-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 69119
    Prev
    Next
    INDEX
    Explanations

    figures and document references

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

    The neuron detects standalone numeric labels and figure or section numbers (e.g. the “8” in “FIG. 8” or the “1” and “2” in section headings) in the patent text.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
     and
    -1.46
    もうすぐ
    -1.35
    明日は
    -1.35
    昨日は
    -1.31
    にします
    -1.26
    今週
    -1.25
     or
    -1.23
    我が家の
    -1.16
    先週
    -1.15
    に使用
    -1.13
    POSITIVE LOGITS
    pography
    1.30
    もあって
    1.30
     parede
    1.27
     岁
    1.26
    并未
    1.23
    iin
    1.21
    ofus
    1.20
     uLocal
    1.20
    teenth
    1.19
    urement
    1.18
    Activations Density 0.006%

    No Known Activations