© 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. 50441
    Prev
    Next
    INDEX
    Explanations

    deaths and injuries

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

    news articles reporting accidents, deaths, or injuries.

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

    The neuron is highly responsive to numeric tokens (ages, counts, statistics, and other digit‐based expressions).

    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
     for
    -1.57
     and
    -1.48
     five
    -1.45
     in
    -1.44
     four
    -1.42
     eight
    -1.40
     more
    -1.40
     three
    -1.38
     two
    -1.38
     when
    -1.37
    POSITIVE LOGITS
     maniere
    1.55
    鵙
    1.40
     conster
    1.35
     materie
    1.30
     terapia
    1.26
    쐈
    1.24
    Provide
    1.23
    튿
    1.23
     zoude
    1.22
     atmosfera
    1.21
    Activations Density 0.010%

    No Known Activations