© 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. 102306
    Prev
    Next
    INDEX
    Explanations

    Olymp and Philipp variants

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

    The neuron is highly responsive to document‐formatting markers—especially runs of “%” used as section dividers or decorative separators.

    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
     People
    -3.50
    ↵
    -3.09
     
    -3.06
     Their
    -2.97
     It
    -2.91
     vorigen
    -2.78
     Coming
    -2.73
    1
    -2.73
     Before
    -2.70
     frühen
    -2.56
    POSITIVE LOGITS
    s
    3.20
    cherichia
    3.20
    ーズン
    2.98
    haviours
    2.98
    ſſen
    2.91
    ዡ
    2.84
    埸
    2.84
    횃
    2.83
    ſſi
    2.83
    itiert
    2.80
    Activations Density 0.007%

    No Known Activations