© 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. 10-GEMMASCOPE-RES-131K
    7. 113817
    Prev
    Next
    INDEX
    Explanations

    years to come

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

    The neuron detects mentions of multi‐year (or multi‐decade) time spans extending into the future (e.g. “years to come”).

    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
     desmotiv
    -1.88
     parachoque
    -1.82
     moeili
    -1.80
    これで
    -1.75
    źródło
    -1.74
    鰊
    -1.73
     sabias
    -1.70
    lepaskan
    -1.69
     pintor
    -1.62
     opgenomen
    -1.61
    POSITIVE LOGITS
     –
    2.38
     be
    2.22
     —
    1.91
     later
    1.84
     latter
    1.73
    ’
    1.59
     
    1.55
    …
    1.53
     />
    1.51
    手下
    1.51
    Activations Density 0.009%

    No Known Activations