© 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. 118115
    Prev
    Next
    INDEX
    Explanations

    standardization and consistency

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

    The neuron detects formal academic “perfect” or passive constructions—especially phrases like “have been …,” “lack of …,” or similar auxiliary+past-participle patterns.

    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
     cómics
    -0.87
     gimnasio
    -0.84
     ترکیب
    -0.82
    人で
    -0.79
    вають
    -0.79
    nais
    -0.79
     würd
    -0.78
    驤
    -0.78
     Kombination
    -0.78
     juguetes
    -0.78
    POSITIVE LOGITS
     uniform
    4.25
     uniformity
    4.09
     standardized
    3.91
     consistency
    3.75
     consistent
    3.67
     standardi
    3.66
     standardization
    3.66
    uniform
    3.58
     standardize
    3.44
     standardised
    3.42
    Activations Density 0.229%

    No Known Activations