الرئيسية المنصة الوثائق الأخبار الأوراق البحثية الدروس مجموعات البيانات الموسوعة SOTA نماذج LLM لوحة الأداء GPU الفعاليات

العربية

Question Answering On Strategyqa

المقاييس

Accuracy

النتائج

نتائج أداء النماذج المختلفة على هذا المعيار القياسي

		Paper Title	Repository
PaLM 2 (few-shot, CoT, SC)	90.4	PaLM 2 Technical Report
Rethinking with retrieval (GPT-3)	77.73	Rethinking with Retrieval: Faithful Large Language Model Inference
Self-Evaluation Guided Decoding (Codex, CoT, single reasoning chain, 6-shot gen, 4-shot eval)	77.2	-	-
U-PaLM 540B	76.6	Transcending Scaling Laws with 0.1% Extra Compute	-
PaLM 540B	76.4	Transcending Scaling Laws with 0.1% Extra Compute	-
Minerva 540B	61.9	Transcending Scaling Laws with 0.1% Extra Compute	-
SearchChain	-	Chain-of-Action: Faithful and Multimodal Question Answering through Large Language Models
Least-to-Most	-	Least-to-Most Prompting Enables Complex Reasoning in Large Language Models
SearchChain	-	Search-in-the-Chain: Interactively Enhancing Large Language Models with Search for Knowledge-intensive Tasks
CoA w/o actions	-	Chain-of-Action: Faithful and Multimodal Question Answering through Large Language Models
CoA	-	Chain-of-Action: Faithful and Multimodal Question Answering through Large Language Models
Least-to-Most	-	Chain-of-Action: Faithful and Multimodal Question Answering through Large Language Models

0 of 12 row(s) selected.