© 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. 130605
    Prev
    Next
    INDEX
    Explanations

    as to or as regards

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

    The neuron responds to short function words—especially the preposition “to” (and other connective prepositions like “as,” “against,” “regard(s),” and parts of “insofar”) that introduce subordinate or legalistic phrases.

    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
     contributes
    -1.59
     Even
    -1.59
     Like
    -1.55
    ”、“
    -1.52
     After
    -1.52
     improves
    -1.51
     invaluable
    -1.51
     considerable
    -1.48
     encourages
    -1.47
    сит
    -1.46
    POSITIVE LOGITS
    扨
    1.86
     algunos
    1.65
     certaines
    1.62
     them
    1.59
     others
    1.50
     alcuni
    1.48
     Apakah
    1.48
    ḩ
    1.39
     alguns
    1.38
    “
    1.38
    Activations Density 0.022%

    No Known Activations