© 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. 34-GEMMASCOPE-RES-131K
    7. 102677
    Prev
    Next
    INDEX
    Explanations

    is or are followed by descriptive words

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

    The neuron strongly activates on forms of the copular/auxiliary “to be” (e.g. is, are, be, been) indicating it’s detecting occurrences of “be” verbs and passive‐style constructions.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
     pojawi
    -1.04
     burada
    -1.04
    Inherited
    -1.01
     présents
    -0.99
    corporated
    -0.98
    imposed
    -0.98
    acije
    -0.96
    േ
    -0.94
    +```
    -0.93
     pojawia
    -0.93
    POSITIVE LOGITS
     دارای
    1.40
     comprised
    1.38
     composed
    1.33
     able
    1.32
     made
    1.10
     equipped
    1.06
     preceded
    1.06
     blessed
    1.03
     voorzien
    1.02
     a
    1.01
    Activations Density 0.233%

    No Known Activations