© 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. 22-GEMMASCOPE-RES-131K
    7. 54275
    Prev
    Next
    INDEX
    Explanations

    recent time references

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

    The main thing this neuron does is spot expressions of recent time (e.g. “yesterday,” “the other day,” “last week,” etc.).

    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
     folyam
    -1.11
    чными
    -1.05
    ächter
    -1.05
    uktion
    -1.04
    requipa
    -1.04
    AsStream
    -1.03
    ValueStyle
    -1.02
     большинстве
    -1.00
    เคย
    -0.99
    cún
    -0.98
    POSITIVE LOGITS
     new
    1.20
     yesterday
    1.19
     recently
    1.10
     today
    1.05
     recent
    1.04
     when
    0.99
    isteren
    0.99
    зацию
    0.98
     at
    0.97
    Exelente
    0.97
    Activations Density 0.030%

    No Known Activations