Unverified50% confidenceSolutionExact time
The authors cancel out the confounding effects of fluency and text quality by reading the same sentence with two different models.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
Unverified挪威斯塔万格大学阅读研究中心的实验表明，阅读纸质文本的受试者在理解测试中表现显著优于阅读同等数字文本的对照组69% similarUnverifiedSummary-type reading apps are fundamentally secondary-processed conclusions that strip away the author's reasoning process, case studies, and thinking paths, which is the root cause of why many people forget what they read right after finishing.69% similarUnverified仅仅改变输入文本的措辞和结构，就能让同一个模型的性能产生数十个百分点的差异68% similarUnverifiedThe bi-encoder model can pre-compute document vectors offline and build an index for fast retrieval, but the two texts cannot see each other during the encoding stage, making it impossible to model fine-grained word-level interactions.67% similarUnverified自编码模型（如BERT系列）通过双向注意力理解上下文，在阅读理解任务上表现优异但不适合开放式生成66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/907912API
curl https://kongchang.com/api/v1/knowledge/claims/907912MCP
get_claim(id=907912)