© 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 EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 47485
    Prev
    Next
    INDEX
    Explanations

    tomorrow and yesterday

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

    The neuron strongly activates on the word “tomorrow” (including its sub‐token parts), i.e. it detects occurrences of “tomorrow.”

    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
    iding
    -0.96
    ρε
    -0.96
     lingue
    -0.94
     patate
    -0.93
    Choosing
    -0.92
    Of
    -0.91
     rosa
    -0.90
    近年来
    -0.88
    จาก
    -0.86
    近期
    -0.86
    POSITIVE LOGITS
     day
    1.95
     morning
    1.91
     Yesterday
    1.68
     Tomorrow
    1.60
    morrow
    1.55
     día
    1.45
     tomorrow
    1.42
     afternoon
    1.42
     night
    1.27
     yesterday
    1.18
    Activations Density 0.011%

    No Known Activations