© 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. 47987
    Prev
    Next
    INDEX
    Explanations

    myeloablative or Eloctate

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

    medical terminology, especially focusing on stem cells, transplantation procedures, and drug names.

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

    The neuron is keyed to the sub‐string “elo” (it fires strongly on tokens like my elo-…, E loctate, melo-die, H eloise, etc.).

    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
    瀲
    -2.50
    狆
    -2.47
    鏨
    -2.45
    
    -2.34
     continúas
    -2.34
    Ꮵ
    -2.33
    笮
    -2.31
     frecuentemente
    -2.30
     图库
    -2.30
    淶
    -2.28
    POSITIVE LOGITS
    ſſen
    2.75
     was
    2.69
     pero
    2.66
    在他
    2.52
    ༷
    2.47
    在她
    2.44
    
    2.42
    “
    2.38
     siguientes
    2.30
     dei
    2.27
    Activations Density 0.031%

    No Known Activations