© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 75343
    Prev
    Next
    INDEX
    Explanations

    medical outcomes and conditions

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

    medical or scientific terms related to patient outcomes, treatments, or research findings.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron fires on long, complex multi‐syllabic words (typically technical or Latinate terms).

    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
    熵
    -1.30
    為に
    -1.16
    Œuvres
    -1.07
    “
    -1.05
     quase
    -1.05
     presque
    -1.05
     uſ
    -1.05
     empêcher
    -1.05
     maniere
    -1.04
     cuatro
    -1.03
    POSITIVE LOGITS
     should
    2.63
     will
    2.34
     may
    1.64
     hopefully
    1.62
     now
    1.54
     would
    1.52
     and
    1.43
     please
    1.43
     can
    1.41
     could
    1.41
    Activations Density 0.027%

    No Known Activations