© 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 EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 10-GEMMASCOPE-RES-131K
    7. 111350
    Prev
    Next
    INDEX
    Explanations

    deleterious effects

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

    technical terms in scientific or engineering documents, particularly related to machine components and negative health effects.

    oai_token-act-pair · claude-3-7-sonnet-20250219Triggered by @neilrathi

    The neuron strongly activates on uncommon, domain-specific technical or scientific jargon words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
    Not in Any Lists

    No Comments

    Negative Logits
     unn
    -2.64
    ☫
    -2.56
    却又
    -2.56
    癰
    -2.44
     duyg
    -2.41
     essayer
    -2.38
    描いて
    -2.36
    戋
    -2.34
    乄
    -2.30
     était
    -2.27
    POSITIVE LOGITS
    t
    2.83
    ↵↵
    2.81
    _
    2.73
    ра
    2.70
                    
    2.41
    I
    2.34
     With
    2.33
    }-\
    2.30
    re
    2.23
     Their
    2.23
    Activations Density 0.020%

    No Known Activations