© 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-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 114156
    Prev
    Next
    INDEX
    Explanations

    instruction verbs and their targets

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

    imperative verbs that instruct the reader to perform an action.

    oai_token-act-pair · claude-4-5-sonnet
    Triggered by @bfgcn7frf5
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Negative Logits
     preparo
    -1.72
    伈
    -1.63
    бе
    -1.57
    cress
    -1.55
    凩
    -1.54
    ୪
    -1.52
    大部分
    -1.52
     awakeFromNib
    -1.49
    
    -1.49
    :✨
    -1.48
    POSITIVE LOGITS
     your
    1.64
     or
    1.64
     to
    1.53
     in
    1.48
    转换成
    1.47
    }
    1.40
    えました
    1.38
     the
    1.37
     gwia
    1.37
     yourself
    1.27
    Activations Density 0.039%

    No Known Activations