© 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. 60987
    Prev
    Next
    INDEX
    Explanations

    even despite or though

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

    This neuron activates on concessive or contrasting adverbs and conjunctions (e.g., “even,” “though,” “despite,” “never”) that introduce a qualification or contrast.

    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
     THERE
    -1.88
    vities
    -1.60
     np
    -1.43
    5
    -1.42
     hig
    -1.41
    llllllll
    -1.41
    8
    -1.39
    ///</
    -1.38
    9
    -1.38
     cb
    -1.38
    POSITIVE LOGITS
     this
    1.89
     many
    1.87
     after
    1.78
     two
    1.63
     during
    1.59
     several
    1.52
     their
    1.52
     those
    1.50
     multifaceted
    1.49
     numerous
    1.46
    Activations Density 0.019%

    No Known Activations