© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 22-GEMMASCOPE-RES-131K
    7. 12562
    Prev
    Next
    INDEX
    Explanations

    potential for harm

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_22/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     fortfarande
    -1.16
    抿
    -1.13
     soixante
    -1.12
     essayer
    -1.11
    咂
    -1.10
     projets
    -1.09
    暗暗
    -1.08
    塒
    -1.07
     comparaison
    -1.07
     quibus
    -1.07
    POSITIVE LOGITS
     both
    1.53
     harm
    1.36
     danger
    1.34
     sowohl
    1.31
     either
    1.29
    Both
    1.26
    (!__
    1.23
     både
    1.21
     entweder
    1.20
     сначала
    1.18
    Activations Density 0.055%

    No Known Activations