The following text is original practice text.
The human brain allocates attention selectively, filtering the vast majority of sensory input before it reaches conscious awareness. This filtering is not passive but actively predictive: the brain generates expectations based on prior experience and attends preferentially to inputs that deviate from those expectations. A dripping faucet noticed only when it stops illustrates the principle. The system is efficient but produces characteristic blind spots, causing humans to miss events that fall exactly within the predicted pattern.
As used in the text, what does "blind spots" most nearly mean?
- AAreas of the visual field where the optic nerve connects to the retina
- BWeaknesses in a person's moral judgment that are caused by emotional attachment
- CKinds of input that the predictive system consistently fails to registerCorrect
- DGaps in scientific knowledge that prevent a theory from being fully tested
Why C is right
In context, 'blind spots' refers to the kinds of input the brain's predictive filtering misses—events that match predicted patterns are not flagged for attention. This is a figurative use meaning systematic failures to perceive.
Why the other 3 choices miss
- A
- Incorrect because this is the literal anatomical meaning of 'blind spot,' but the text uses the term figuratively in a cognitive context.
- B
- Incorrect because moral judgment and emotional attachment are not discussed; the 'blind spots' are perceptual, not moral.
- D
- Incorrect because the text is about perception, not about gaps in scientific theory.
The following text is original practice text.
For decades, researchers claimed that bilingualism confers a measurable 'cognitive advantage'—that managing two languages simultaneously improves executive functions such as attention, cognitive flexibility, and inhibitory control. This 'bilingual advantage' hypothesis attracted widespread popular attention. However, a wave of replication studies beginning in the 2010s failed to reproduce the original findings, and several meta-analyses found that the effect sizes reported in initial studies were substantially smaller or absent when larger, more representative samples were used. The scientific community now regards the bilingual advantage as contested at best, with ongoing debate about whether methodological inconsistencies across studies have masked a real but small effect or whether the effect does not exist at all.
Which choice best states the main idea of the text?
- AThe bilingual cognitive advantage hypothesis, once widely accepted, has been substantially undermined by replication failures and remains a contested claim.Correct
- BThe bilingual advantage is a robust finding that has been confirmed across multiple languages and cultural contexts, though the effect is smaller than initially reported.
- CResearchers should abandon the study of bilingualism's cognitive effects because the methodological challenges of this research area make reliable findings impossible.
- DBilingualism definitively does not provide any cognitive advantage, as multiple large-scale studies have proven the original findings were the result of methodological errors.
Why A is right
The passage describes the rise of the bilingual advantage hypothesis, then its challenge from replication failures and meta-analyses, concluding that it is 'contested at best.' Choice A accurately captures this trajectory—a once widely accepted claim now substantially undermined and contested—without overstating the negative evidence.
Why the other 3 choices miss
- B
- Incorrect because the text describes replication failures and meta-analyses showing smaller or absent effects—not that the advantage has been confirmed across contexts.
- C
- Incorrect because the text describes ongoing debate, not a conclusion that the research area should be abandoned.
- D
- Incorrect because the text does not say bilingualism 'definitively' provides no advantage; it says the effect is contested and may be small rather than nonexistent.
The following text is original practice text.
A long-standing debate in artificial intelligence research concerns whether current large language models (LLMs) genuinely understand the text they process or merely learn to produce statistically plausible outputs without any underlying comprehension. Critics of comprehension claims point to the fact that LLMs sometimes make confident errors on simple logical or factual tasks that any competent human reader would handle correctly. Proponents counter that 'understanding' is a poorly defined concept even in human cognition, and that the performance differences between LLMs and humans on such tasks do not establish that LLMs lack a form of understanding—only that their processing differs from human processing in ways that sometimes produce errors.
Which choice best states the main idea of the text?
- AWhether large language models genuinely understand language is an open debate, with critics citing errors and proponents questioning whether understanding is well-defined.Correct
- BThe philosophy of AI has conclusively shown that there is no meaningful distinction between genuine understanding and the production of statistically plausible outputs.
- CLarge language models produce statistically plausible text by processing patterns in training data, a method fundamentally different from and inferior to human language comprehension.
- DCurrent large language models do not genuinely understand text, as demonstrated by their consistent failure to handle the simple logical tasks that characterize true comprehension.
Why A is right
The passage presents both sides of an ongoing debate: critics point to errors as evidence against comprehension; proponents argue 'understanding' is ill-defined and error patterns don't settle the question. Choice A captures the contested, unresolved nature of this debate.
Why the other 3 choices miss
- B
- Incorrect because the text presents this debate as ongoing and unresolved, not conclusively settled in either direction.
- C
- Incorrect because the text does not endorse the claim that LLMs' method is 'inferior'—it presents both views without resolving which is correct.
- D
- Incorrect because the text presents this as critics' view, not as an established conclusion, and notes proponents dispute this interpretation.