© 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. 22-GEMMASCOPE-RES-131K
    7. 26855
    Prev
    Next
    INDEX
    Explanations

    extreme defects

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

    The neuron is primarily detecting numeric tokens and numerical expressions (e.g. standalone numbers, measurement values, ratings, and related symbols).

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    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
     légèrement
    -1.09
     bättre
    -0.97
    ţele
    -0.95
    自体は
    -0.90
    hésitez
    -0.90
    Imagem
    -0.90
    氿
    -0.86
    ണ്
    -0.85
     după
    -0.84
    こっち
    -0.84
    POSITIVE LOGITS
     completely
    1.76
     totally
    1.48
     entirely
    1.43
     Completely
    1.36
     even
    1.35
     extreme
    1.27
     outright
    1.20
     fully
    1.16
     all
    1.09
     полностью
    1.08
    Activations Density 0.107%

    No Known Activations