© 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. 65170
    Prev
    Next
    INDEX
    Explanations

    identifies entities as

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

    This neuron activates on the copular “is” in entity‐definition sentences (e.g. the “[Subject] is a …” patterns in encyclopedic lead lines).

    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
     and
    -1.42
    ":"",
    -1.40
    )));
    -1.32
    "};
    -1.30
    )}=
    -1.26
    '){
    -1.21
     then
    -1.20
    '){
    
    -1.20
    '}
    -1.16
    ')));
    -1.16
    POSITIVE LOGITS
     cuando
    1.46
    トッピング
    1.40
     amikor
    1.33
     любых
    1.32
     recommande
    1.27
     you
    1.26
     étranger
    1.26
    怎么能
    1.23
    谁能
    1.21
     ketika
    1.20
    Activations Density 0.746%

    No Known Activations