© 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. 22-GEMMASCOPE-RES-131K
    7. 35836
    Prev
    Next
    INDEX
    Explanations

    drive towards improvement

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

    This neuron fires on purpose-oriented function words that introduce goals or objectives—especially the infinitival “to” and similar direction markers like “for” or “towards.”

    oai_token-act-pair · o4-miniTriggered 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
    Not in Any Lists

    No Comments

    Negative Logits
     nuevos
    -1.30
     daardoor
    -1.28
     vrijdag
    -1.22
     Pueden
    -1.20
     kostenlosen
    -1.18
     leeren
    -1.16
     Incluye
    -1.16
     daarbij
    -1.16
     Actualmente
    -1.16
     permitió
    -1.16
    POSITIVE LOGITS
     better
    1.21
     and
    1.15
     (
    1.13
     all
    1.13
     Catawiki
    1.03
     pulumi
    1.00
     its
    0.98
     which
    0.98
     global
    0.94
    mín
    0.91
    Activations Density 0.038%

    No Known Activations