© 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. 34-GEMMASCOPE-RES-131K
    4. 25957
    Prev
    Next
    INDEX
    Explanations

    dates and days or technical terms

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

    The neuron fires on numeric tokens, especially when they’re part of section or document identifiers (e.g. headings like “ISA 230” or other numbered codes).

    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
    此前
    -0.77
    гно
    -0.76
    恶意
    -0.75
    一口气
    -0.74
    飾
    -0.72
     jueves
    -0.69
    マネ
    -0.68
     mehrerer
    -0.68
     Đây
    -0.67
    OPIN
    -0.67
    POSITIVE LOGITS
     vouch
    1.01
    Recon
    0.96
    Agree
    0.91
     fysis
    0.86
    原始
    0.84
     Comput
    0.83
     segregation
    0.82
     procédures
    0.82
    hålla
    0.81
     Procedures
    0.81
    Activations Density 0.012%

    No Known Activations