© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 32824
    Prev
    Next
    INDEX
    Explanations

    be or to be done

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

    The neuron is detecting high‐frequency auxiliary and function words (like “have,” “be,” “will,” “to,” “gone on,” “never,” etc.) that form verb‐phrase constructions indicating tense or aspect.

    oai_token-act-pair · o4-mini
    Triggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     aset
    -1.30
    матика
    -1.21
     kram
    -1.19
     seks
    -1.17
     itens
    -1.16
     Polonia
    -1.16
     Kelas
    -1.16
     Inscrivez
    -1.15
     todav
    -1.14
     peper
    -1.14
    POSITIVE LOGITS
     happening
    1.48
     happen
    1.47
     that
    1.39
     there
    1.35
     happened
    1.30
     kteří
    1.21
     a
    1.21
     time
    1.20
     different
    1.20
     asociaciones
    1.16
    Activations Density 0.040%

    No Known Activations