© 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. 10-GEMMASCOPE-RES-131K
    4. 86921
    Prev
    Next
    INDEX
    Explanations

    premature death and infants

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

    The neuron detects the word "premature" when used in medical and scientific contexts, particularly when associated with concepts like death, mortality, birth, or disease.

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

    the neuron consistently activates on numerical mentions of “premature” (e.g. premature death, premature mortality, premature infants) and their associated statistics.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     ie
    -2.41
    ]
    -2.28
     ]
    -2.19
    \
    -2.16
    休日
    -2.11
     owns
    -2.09
    一根
    -2.08
     has
    -2.08
     celebrates
    -2.05
     pero
    -2.00
    POSITIVE LOGITS
    酲
    2.80
    ਗ਼
    2.73
    ⢓
    2.59
    囖
    2.45
     laad
    2.44
    ཋ
    2.44
    阄
    2.41
     oman
    2.38
    Здесь
    2.33
    趼
    2.31
    Activations Density 0.005%

    No Known Activations