© 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
    5. 31-GEMMASCOPE-2-RES-262K
    7. 129458
    Prev
    Next
    INDEX
    Explanations

    I know what

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

    surprising negative observations

    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_31_width_262k_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
     was
    0.40
     newly
    0.38
     rented
    0.37
     started
    0.37
     amounted
    0.37
     cities
    0.36
     hilly
    0.36
    ↵↵↵↵↵↵↵
    0.35
    ↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵↵
    0.35
     large
    0.35
    POSITIVE LOGITS
    regarding
    0.49
    能否
    0.45
    能不能
    0.44
    Nevertheless
    0.43
    não
    0.43
    нимание
    0.43
    Surprisingly
    0.43
    ไม่มี
    0.43
     בפ
    0.43
    surprisingly
    0.42
    Activations Density 0.000%

    No Known Activations