© 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. 10-GEMMASCOPE-RES-131K
    4. 21940
    Prev
    Next
    INDEX
    Explanations

    lists (health, science, technical)

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

    instances where the word "and" connects items in lists or phrases in scientific/technical documents.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron specializes in detecting numeric tokens—any digits, ordinal markers, or other number-like sequences within the 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_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
    +]
    -1.25
    德
    -1.24
     家族
    -1.22
     絶
    -1.21
    、
    -1.20
    nend
    -1.19
    {
    
    -1.16
    内
    -1.16
    ア
    -1.14
    𝑖
    -1.14
    POSITIVE LOGITS
    Воспитатель
    1.28
     a
    1.27
     понравился
    1.26
     where
    1.23
    This
    1.21
     by
    1.20
    摻
    1.19
     бампер
    1.13
     Ihren
    1.12
     вовсе
    1.10
    Activations Density 0.225%

    No Known Activations