© 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
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 20423
    Prev
    Next
    INDEX
    Explanations

    expressions involving variables

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

    This neuron responds to cue words and phrases that structure math problem statements—e.g. “Let,” “where,” “assuming,” “What is,” “Express,” “Determine,” etc.—marking definitions, conditions, and question prompts.

    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
     amd
    -1.80
     harika
    -1.77
     ücretsiz
    -1.63
    ܊
    -1.63
    
    -1.55
     cea
    -1.54
    hoge
    -1.53
     astonishing
    -1.53
     basit
    -1.52
    những
    -1.51
    POSITIVE LOGITS
     because
    2.14
     when
    1.80
     Both
    1.73
     both
    1.71
     Generally
    1.69
     Normally
    1.69
     After
    1.65
     Before
    1.63
     Initially
    1.62
     When
    1.60
    Activations Density 0.316%

    No Known Activations