© 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. 22-GEMMASCOPE-RES-131K
    7. 25483
    Prev
    Next
    INDEX
    Explanations

    listing multiple options

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

    The neuron fires strongly on opening parenthesis characters, i.e. it detects parenthetical clauses.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/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
    -
    -0.98
    BeginContext
    -0.98
     Baths
    -0.88
    艙
    -0.85
    kl
    -0.84
    '
    -0.84
     ganglion
    -0.83
    quarters
    -0.82
     Exploit
    -0.81
     überrascht
    -0.81
    POSITIVE LOGITS
     दोनों
    1.05
     sekä
    1.03
     båda
    0.96
    両方
    0.95
     olmak
    0.95
    瘢
    0.94
     รวม
    0.93
    どちらも
    0.92
     keduanya
    0.90
     can
    0.88
    Activations Density 0.162%

    No Known Activations