© 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. 44420
    Prev
    Next
    INDEX
    Explanations

    IDs, PCs, CDs

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

    The neuron fires on uppercase initialisms (like “IDs,” “PCs,” “CDs”)—i.e. acronyms/abbreviations in all caps.

    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
    ſelves
    -2.88
    姈
    -2.84
     ;
    -2.69
    ༀ
    -2.59
     are
    -2.58
    リコン
    -2.44
    ⟞
    -2.42
     :
    -2.38
    erintah
    -2.33
    rozco
    -2.25
    POSITIVE LOGITS
    筇
    2.52
    閪
    2.42
     participé
    2.41
    '
    2.38
    〮
    2.34
    〰
    2.30
    锃
    2.27
    薷
    2.27
    塒
    2.27
     câbles
    2.25
    Activations Density 0.012%

    No Known Activations