© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-3-27B-IT
    5. 31-GEMMASCOPE-2-RES-262K
    7. 9165
    Prev
    Next
    INDEX
    Explanations

    Depressed and other phrases

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

    instructions that define or emphasize a persona/style or behavior (often in quoted or capitalized form and with roleplay cues).

    oai_token-act-pair · gpt-5
    Triggered by @vetterc0
    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
    Negative Logits
    based
    0.41
     significantly
    0.41
     related
    0.40
     based
    0.39
     significant
    0.38
     use
    0.36
     using
    0.35
     comprise
    0.35
    trong
    0.35
     specific
    0.34
    POSITIVE LOGITS
     matahari
    0.38
     प्यार
    0.34
     coração
    0.34
     ساعة
    0.34
    Liverpool
    0.33
     улы
    0.33
     kisah
    0.33
    ﺷ
    0.33
     жизни
    0.33
     жизнь
    0.33
    Activations Density 0.157%

    No Known Activations