© 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. 11985
    Prev
    Next
    INDEX
    Explanations

    thresholds or conditions

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

    The neuron detects tokens associated with numerical specifications and quantitative measurements (e.g. thresholds, concentrations, durations, exposure levels).

    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
     スリム
    -0.99
     マフラー
    -0.96
     pomme
    -0.91
    драт
    -0.91
     Prueba
    -0.90
     patate
    -0.90
     enorm
    -0.90
     APPROPRI
    -0.90
    衚
    -0.89
     occuper
    -0.88
    POSITIVE LOGITS
     begin
    1.58
     begins
    1.49
     sustainable
    1.30
     Begin
    1.27
     acceptable
    1.27
     mulai
    1.23
     beginning
    1.22
     onset
    1.20
     comfortable
    1.20
     where
    1.20
    Activations Density 0.117%

    No Known Activations