© 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
    2. Gemma-2-27B
    3. 22-GEMMASCOPE-RES-131K
    4. 79882
    Prev
    Next
    INDEX
    Explanations

    interval timing

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

    The main thing this neuron does is detect the term “interval” (including its plural “intervals”) in text.

    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
     unending
    -2.86
     exceedingly
    -2.78
     overarching
    -2.72
    ’,
    -2.56
     immense
    -2.55
     wondrous
    -2.55
     tentando
    -2.45
    ほとん
    -2.39
     kleinere
    -2.38
     insanely
    -2.36
    POSITIVE LOGITS
    d
    3.02
     After
    2.88
     For
    2.77
     Then
    2.59
    кому
    2.53
     .....
    2.50
     после
    2.42
    ၅
    2.42
     revital
    2.39
     bestäm
    2.39
    Activations Density 0.005%

    No Known Activations