© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B
    5. 53-GEMMASCOPE-2-RES-65K
    7. 17555
    Prev
    Next
    INDEX
    Explanations

    unintended consequences and errors

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

    prevents errors and unexpected outcomes

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-27b-pt/resid_post/layer_53_width_65k_l0_medium
    Prompts (Dashboard)
    392,802 prompts, 256 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Minimize
    0.49
     Minim
    0.47
     Developing
    0.45
    ปรับ
    0.45
     Primarily
    0.45
     minimise
    0.45
     minimising
    0.42
    Developing
    0.42
    Mechanical
    0.42
     Relatively
    0.41
    POSITIVE LOGITS
     unintentionally
    0.89
     incorrectly
    0.88
     crashing
    0.87
     messed
    0.87
     crashes
    0.84
     unexpectedly
    0.83
     inadvertently
    0.83
     crash
    0.82
     unexpected
    0.82
     unintended
    0.82
    Activations Density 0.102%

    No Known Activations