© 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. 34-GEMMASCOPE-RES-131K
    7. 33339
    Prev
    Next
    INDEX
    Explanations

    tolerate or not tolerate

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

    The neuron responds to negative or prohibitive constructions—words and phrases expressing negation, prohibition, or condemnation (e.g. “no,” “not,” “shouldn’t,” “don’t,” “hate,” “condemning”).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/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
    Tó
    -0.78
    性に
    -0.73
     vermeiden
    -0.73
     remarqu
    -0.73
     compareTo
    -0.72
    дневно
    -0.72
    桎
    -0.72
     tyd
    -0.72
    kø
    -0.72
     axial
    -0.71
    POSITIVE LOGITS
     con
    2.75
     approve
    2.13
     sanction
    2.11
     endorse
    2.03
     cond
    1.91
     approval
    1.85
     approving
    1.77
     encourage
    1.76
     tolerate
    1.75
     toler
    1.74
    Activations Density 0.060%

    No Known Activations