© 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. 90795
    Prev
    Next
    INDEX
    Explanations

    tailoring advice and job descriptions

    np_acts-logits-general · gemini-2.5-flash-lite
    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
    !
    0.52
    а
    0.49
    і
    0.46
    br
    0.45
    q
    0.43
    f
    0.43
     visually
    0.43
    一
    0.43
    и
    0.42
    features
    0.40
    POSITIVE LOGITS
     disgraceful
    0.54
     conséquences
    0.51
     baisse
    0.47
     gått
    0.47
     engend
    0.46
     coû
    0.45
     passada
    0.45
     testemun
    0.44
     retirada
    0.44
     persegu
    0.44
    Activations Density 0.002%

    No Known Activations