© 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. 107741
    Prev
    Next
    INDEX
    Explanations

    future and recent topics

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

    The neuron fires on the subtitle portion of scientific paper titles—especially on the colon and the coordinating conjunction “and” (and the key words linked by it) that join two title elements.

    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
     simpel
    -1.16
     gede
    -1.15
     vokal
    -1.12
    的一部分
    -1.12
     човек
    -1.11
     berge
    -1.11
     externes
    -1.07
     périodes
    -1.06
    éroport
    -1.05
     menyen
    -1.05
    POSITIVE LOGITS
     challenges
    2.77
     future
    2.28
     current
    2.27
     trends
    2.19
     opportunities
    2.14
     prospects
    2.14
     recent
    1.96
     potential
    1.90
     possible
    1.84
     problems
    1.78
    Activations Density 0.124%

    No Known Activations