© 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. 98456
    Prev
    Next
    INDEX
    Explanations

    difference comparing properties

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

    The neuron responds to explicit quantitative statements—numbers (and their surrounding count‐words like “variables,” “in,” “is,” etc.) conveying a numeric fact.

    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
    日在
    -0.83
     implore
    -0.82
     Backend
    -0.81
    لے
    -0.81
    っていると
    -0.77
     expos
    -0.77
     senderos
    -0.76
    нитель
    -0.76
    viembre
    -0.75
    ünden
    -0.75
    POSITIVE LOGITS
     unlike
    1.53
    unlike
    1.37
     Unlike
    1.23
     differs
    1.20
    Unlike
    1.19
     distinguishes
    1.13
     resembles
    1.10
     similarities
    1.09
     therefore
    1.06
     отличие
    1.05
    Activations Density 0.122%

    No Known Activations