© 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. 19115
    Prev
    Next
    INDEX
    Explanations

    foreign language words

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

    The neuron strongly activates on very common, high-frequency verbs (especially forms of “be,” “have,” and other frequent auxiliaries or past participles).

    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
     it
    -1.73
     in
    -1.67
     at
    -1.55
    There
    -1.53
     after
    -1.46
    there
    -1.45
    AILABILITY
    -1.37
    After
    -1.33
    yó
    -1.33
    POSITIVE LOGITS
    我一直
    1.50
    <bos>
    1.50
    比较好
    1.41
     douane
    1.34
     котором
    1.32
     horloge
    1.32
     (~(
    1.31
     diverses
    1.30
     rigue
    1.30
     Palabra
    1.29
    Activations Density 0.069%

    No Known Activations