© 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. 73469
    Prev
    Next
    INDEX
    Explanations

    exceptions or variations

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

    This neuron fires on formal prepositional‐phrase constructions that introduce constraints or exclusions—e.g. “minus the empty set” or “subject to the limitations in the disclaimer below.”

    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
     both
    -1.23
    both
    -1.01
    せっかく
    -0.97
     wohnung
    -0.95
     thus
    -0.94
    かなぁ
    -0.93
     zowel
    -0.90
    ريقة
    -0.89
    ऎ
    -0.89
     schlaf
    -0.85
    POSITIVE LOGITS
     oczywiście
    1.27
     natürlich
    1.16
     quelques
    1.15
     конечно
    1.13
    を除
    1.06
     البته
    1.02
     claro
    1.02
     exceptions
    1.01
     неболь
    0.96
    szön
    0.93
    Activations Density 0.083%

    No Known Activations