© 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. 10-GEMMASCOPE-RES-131K
    7. 105374
    Prev
    Next
    INDEX
    Explanations

    last followed by past state or event

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

    The neuron activates on occurrences of the word “last,” marking references to a previous time or prior event.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_10/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
     abbildung
    -1.83
     eerste
    -1.61
     первом
    -1.45
    création
    -1.40
     pierwszym
    -1.38
    鉨
    -1.38
     Après
    -1.35
    zló
    -1.34
    etsk
    -1.34
    tifs
    -1.33
    POSITIVE LOGITS
     a
    1.95
     in
    1.66
    d
    1.36
    ↵
    1.34
     manifests
    1.30
     for
    1.28
     before
    1.26
    section
    1.26
    ta
    1.23
     roligt
    1.21
    Activations Density 0.025%

    No Known Activations