गाइड से जुड़ी प्रश्नोत्तरी · मध्यम स्तर

बेंचमार्क संदूषण प्रश्नोत्तरी

परीक्षण करता है कि बेंचमार्क आइटम प्रशिक्षण डेटा में कैसे लीक होते हैं, एन-ग्राम ओवरलैप और मिन-के% प्रोब जैसे पता लगाने के तरीके, और कैनरी और टाइम स्प्लिट जैसे शमन।

संबंधित मार्गदर्शक पथबेंचमार्क संदूषण
प्रश्न 1 का 8

बेंचमार्क संदूषण क्या है?

प्रत्येक प्रश्न का पूर्वावलोकन करें8 प्रश्न

जैसे ही आप प्रश्नोत्तरी लेते हैं उत्तर और स्पष्टीकरण दिखाई देते हैं।

  1. बेंचमार्क संदूषण क्या है?
    • When a benchmark contains incorrect answer keys
    • When a model is tested on too few questions
    • When benchmark test items or their answers appear in a model's training data
    • When two benchmarks measure the same skill
  2. बिग-बेंच कार्यों में शामिल कैनरी स्ट्रिंग का उद्देश्य क्या है?
    • To let data teams filter those documents from training data and test whether a model saw them
    • To encrypt the answers
    • To mark which tasks are the hardest
    • To identify which model generated each answer
  3. GSM1k अध्ययन ने क्या सुझाव दिया?
    • All models scored higher on the new problems
    • GSM8K has no errors
    • Contamination cannot be measured without training data
    • Some models scored noticeably lower on fresh, matched problems than on GSM8K
  4. GPT-3 पेपर में कौन सा ओवरलैप विश्लेषण चला?
    • Embedding similarity on full documents
    • 13-gram overlap between benchmarks and training data
    • Human review of every training document
    • Min-K% Prob on each benchmark
  5. मिन-के% प्रोब का मूल विचार क्या है?
    • Removing the least common k percent of words from training data
    • Training a second model to classify contaminated text
    • Text the model has seen tends to have few very low-probability tokens
    • Measuring the length of the model's answers
  6. How does a time-based split, as in LiveCodeBench, help detect contamination?
    • It compares performance on problems published before and after the model's training cutoff
    • It measures how long the model takes to answer
    • It splits the model into older and newer layers
    • It only counts problems answered within a time limit
  7. Why does finding no exact overlap not prove a benchmark is clean?
    • Exact matching is too slow to run
    • Benchmarks change their answers over time
    • Models never memorize exact text
    • Paraphrased or translated copies can slip past string matching
  8. In n-gram decontamination, what happens if n is chosen too small?
    • Near-copies with small edits are missed
    • Many false positives from common phrases are flagged
    • The filter stops working entirely
    • Only answers, not questions, are matched