© 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-3-27B-IT
    5. 31-GEMMASCOPE-2-RES-262K
    7. 8163
    Prev
    Next
    INDEX
    Explanations

    AI assistant

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

    This neuron spots the assistant’s self-descriptive policy-and-safety statements, especially “I cannot/absolutely cannot” refusals based on its programming constraints.

    oai_token-act-pair · o4-miniTriggered by @yooniel31
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-it/resid_post/layer_31_width_262k_l0_medium
    Prompts (Dashboard)
    238,145 prompts, 512 tokens each
    Dataset (Dashboard)
    lmsys + oasst1
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     würde
    0.38
     või
    0.37
     或
    0.36
     algum
    0.36
    would
    0.35
    もしくは
    0.35
     poderia
    0.34
     ataupun
    0.34
    就可以
    0.33
     would
    0.33
    POSITIVE LOGITS
    あくまで
    0.54
     භ
    0.36
     narzęd
    0.35
     teknoloj
    0.33
     computadora
    0.33
    TECHN
    0.33
     fundamentally
    0.33
     proizvoda
    0.33
    기술
    0.32
     inanimate
    0.32
    Activations Density 0.579%

    No Known Activations