© 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. 322
    Prev
    Next
    INDEX
    Explanations

    generated by Django

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

    The neuron activates on numeric tokens representing dates—especially four-digit year strings in comments or metadata.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408

    date/time stamps in technical text, especially numeric dates and times in metadata or “Generated on/Date” headers with hyphen/colon-separated sequences.

    oai_token-act-pair · gpt-5Triggered 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
    felizmente
    -1.09
     jaren
    -1.07
     years
    -0.98
     ganó
    -0.97
     thousands
    -0.97
     vuonna
    -0.96
     vagy
    -0.95
    ipedi
    -0.94
    lijker
    -0.94
    ㎍
    -0.94
    POSITIVE LOGITS
     graden
    0.95
     tull
    0.87
     мост
    0.82
     注文
    0.80
     ankara
    0.80
     asos
    0.79
     hant
    0.79
    ϯ
    0.78
    文獻
    0.77
     dispatcher
    0.77
    Activations Density 0.008%

    No Known Activations