© 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. 10-GEMMASCOPE-RES-131K
    7. 119629
    Prev
    Next
    INDEX
    Explanations

    archaic verbs and hematopoiesis

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

    Old English or archaic verb forms like "hath", "doth", and words related to hematopoietic systems.

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

    This neuron detects archaic English verb forms ending in “-th” or “-st,” such as hath, doth, canst, and hast.

    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
    
    -2.81
    ୌ
    -2.72
    怂
    -2.72
    蒹
    -2.63
    脔
    -2.61
     allow
    -2.48
    媗
    -2.47
    轱
    -2.45
    liśmy
    -2.42
    ógicos
    -2.41
    POSITIVE LOGITS
    s
    2.89
     REALLY
    2.67
    即使
    2.44
    (
    2.41
    ſu
    2.41
    很好的
    2.38
    y
    2.34
    想
    2.33
     voleva
    2.28
    可能是
    2.25
    Activations Density 0.001%

    No Known Activations