© 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. 70967
    Prev
    Next
    INDEX
    Explanations

    lessons learned

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

    The neuron detects expressions referring to lessons or takeaways (e.g. “lesson,” “learned,” “teach,” “teaches”).

    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
    ⽬
    -1.18
    雰
    -1.14
    auen
    -1.13
    ligente
    -1.12
    توضیحات
    -1.09
    titue
    -1.07
    几种
    -1.05
    躂
    -1.05
    ЗА
    -1.03
     religiosas
    -1.03
    POSITIVE LOGITS
     from
    2.64
     lessons
    2.06
     that
    1.94
    lessons
    1.65
     lesson
    1.64
    Lessons
    1.38
     about
    1.38
     every
    1.30
     Lessons
    1.30
     warns
    1.30
    Activations Density 0.014%

    No Known Activations