© 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-3-4B-IT
    5. 29-GEMMASCOPE-2-RES-16K
    7. 2260
    Prev
    Next
    INDEX
    Explanations

    negations of exactly, really

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-2-4b-it/resid_post/layer_29_width_16k_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
    包括
    0.87
     WITHOUT
    0.84
     due
    0.83
     without
    0.81
    Due
    0.81
    同时也
    0.80
     +=
    0.80
    ക്കുകയും
    0.79
    )}+\
    0.79
    Without
    0.79
    POSITIVE LOGITS
     really
    2.06
    really
    1.99
     Really
    1.95
    Really
    1.78
     vraiment
    1.47
     exactly
    1.46
    exactly
    1.42
     exactamente
    1.34
     Exactly
    1.30
     exatamente
    1.25
    Activations Density 0.178%

    No Known Activations