© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 100408
    Prev
    Next
    INDEX
    Explanations

    HMS followed by names or insecticide

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

    abbreviations and titles related to military and naval vessels (particularly HMS, Aur/HMS).

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

    The neuron is activated by rare or specialized tokens—uncommon proper names, technical terms, or non-Latin script characters.

    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
    ’
    -3.55
    酲
    -3.41
    ⢓
    -2.84
    ൎ
    -2.77
    觏
    -2.75
    洐
    -2.72
    0
    -2.69
    泔
    -2.59
    镦
    -2.56
     intensely
    -2.50
    POSITIVE LOGITS
    翃
    3.05
     таких
    2.72
    廰
    2.66
    厹
    2.64
    <th>
    2.63
    刄
    2.63
    秶
    2.59
    没有
    2.59
    iyaki
    2.55
    你有
    2.52
    Activations Density 0.013%

    No Known Activations