© 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. 34-GEMMASCOPE-RES-131K
    7. 33566
    Prev
    Next
    INDEX
    Explanations

    lack of controls or audit

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

    This neuron responds to colon characters (":"), especially in time stamps or date/time headers.

    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
    tsv
    -0.95
     werde
    -0.84
     yere
    -0.78
     Constitucional
    -0.76
     you
    -0.76
     syarat
    -0.75
     menyenangkan
    -0.75
     werden
    -0.75
     will
    -0.75
    zuma
    -0.75
    POSITIVE LOGITS
     lacked
    1.20
     lack
    1.06
     audit
    1.03
     reconcili
    1.02
     audi
    1.02
     auditors
    0.98
     controls
    0.95
    存在
    0.94
     impre
    0.94
     risk
    0.93
    Activations Density 0.007%

    No Known Activations