高考风向预测 · 外刊精读 06:Your Brain Doesn't Start from Zero|为什么大脑判断＂现在＂,总要参考＂刚才＂?

MIT 原报道官方配图：画面中橙黄色区域是连接 lateral posterior thalamus（LP）和 anterior cingulate cortex（ACC）的神经轴突。

Researchers identified a brain circuit that helps compare current sensory information with the immediate past. Image: Ning Leow / MIT

Imagine that you are watching a group of dots moving across a screen. Most of them appear to be moving to the left, so you decide that "left" is the correct answer. A few seconds later, another group of dots appears. How does your brain decide what direction they are moving this time?

It might seem that the brain simply examines the new information and makes a fresh decision. But new research from MIT suggests that the process is more complicated. Rather than evaluating every new event from scratch, the brain also compares what is happening now with what happened only moments before.

Researchers have identified a brain circuit that appears to play an important role in this comparison. The circuit connects a region of the thalamus involved in processing sensory information with part of the frontal cortex involved in decision-making.

To understand how the circuit works, scientists trained mice to perform a visual task. The animals watched dots moving on a screen. Although the dots moved in different directions, a certain percentage of them moved together either to the left or to the right. The mice had to decide which direction was dominant.

The difficulty of the task could be changed by adjusting the percentage of dots moving together. For example, when 64 percent of the dots moved left, the answer was relatively clear. When only a small percentage moved together, the choice became more difficult.

Importantly, the researchers also changed the stimulus from one trial to the next. This allowed them to study whether the mice judged each new image independently or whether their most recent experience influenced the next decision.

The results showed that the past mattered. If a mouse had made a correct choice and the next visual signal was similar to the previous one, it was more likely to repeat its answer. If the new signal was very different, however, the mouse was more likely to change its choice.

Interestingly, the pattern changed after an incorrect answer. When a mouse had guessed wrongly, it became less likely to repeat the same decision when the next signal was similar.

Behavior alone, however, could not prove which brain circuit was responsible. The researchers therefore examined a pathway connecting two brain areas. In mice, one of these areas is called the lateral posterior thalamus, or LP. It sends information to an area of the frontal cortex known as the anterior cingulate cortex, or ACC.

Using a two-photon microscope, the scientists recorded neural activity while the mice performed the visual task. Their results suggested that the two regions were doing different jobs. The LP did not appear to make the final decision. Instead, it seemed to report how different the current sensory information was from the recent past. The ACC then used that comparison while forming the animal's choice.

The scientists went one step further. They used a technique called optogenetics to artificially change activity in the pathway between the two brain areas. When they disturbed the circuit, the mice became worse at using recent sensory history to judge the new information. This was important because it provided evidence that the circuit was not simply active during the comparison process — it was actually involved in causing it.

In simple terms, the system appears to work like an adviser. If the animal has just made the correct choice and very little has changed, there may be little reason to change its mind. But if the new information is very different, the brain may need to reconsider.

Such a system could make decision-making more efficient. In a fast-changing world, starting every judgment completely from zero would require the brain to repeatedly process information it had only just encountered.

The research was conducted in mice, so scientists must be careful when applying the findings directly to humans. However, humans have closely related brain structures, and researchers believe similar processes may help people compare present information with recent experience.

The study therefore suggests that perception is not simply about what is in front of us at this exact moment. Even when we think we are judging the present, the immediate past may already be shaping what the present means.

很容易把文章概括成 The study is about memory——不够准确。它真正研究的是：How recent sensory history affects current decisions。也就是：现在的信息 vs 刚刚的信息，大脑进行比较以后，再决定 maintain the previous choice 还是 update the decision。全文结构可以压缩成：

这就是整篇文章的主线。

研究者让小鼠看屏幕上的移动圆点。例如 64% of the dots moved left，那么 left 就是相对明显的趋势；下一次可能变成 16% moved right。于是研究人员不仅可以研究 current direction，还可以研究：How different is the current stimulus from the previous one?

这一步非常关键。因为他们真正想测的不是小鼠会不会看圆点，而是：小鼠有没有利用上一轮的信息来判断这一轮。MIT 报道明确指出，研究人员通过不断改变连续两次任务之间的方向和一致程度，观察小鼠如何根据最近的视觉历史调整判断。

这是全文最容易出一道"为什么"的地方。研究人员不是只 looked at which neurons became active，他们还人为改变 LP → ACC pathway，然后发现：小鼠利用 recent sensory history 的方式被打乱了。所以研究人员得到的证据更强。

MIT 报道中，研究者明确指出，这说明该通路在比较过程中具有 causal involvement，而不是单纯"携带了一些相关信息"。

研究结果显示：LP 更像 comparison adviser——它告诉前额叶"现在和刚才差了多少"；而 ACC 更像 decision maker——它利用这些信息形成最终选择。MIT 研究者的解释是，LP 并不直接做出决定，而是把基于历史的感觉比较信息发送到 ACC，由 ACC 把这些信息转化为与最终选择相关的神经活动。

原报道里有一句特别适合高考：The brain does not evaluate each new event from scratch. 什么意思？当然不是"从抓痕开始"。根据整篇文章：如果大脑使用 previous information，那就说明它不是每次都 start completely new without using what came before。所以 from scratch = from the very beginning, without using previous information——即"从零开始"。

文章最后说：这种系统可能让 decision-making 更加高效。为什么？因为如果环境和刚才几乎一样，大脑没有必要 completely rebuild its understanding，而可以利用 recent history；但是如果变化明显，它又需要 update。所以这个机制本质上是在平衡 stability 和 flexibility。

还是不列长单词表，这一篇留这些即可：

这篇不需要讲脑解剖，更不用要求学生记 lateral posterior thalamus 和 anterior cingulate cortex——可以直接写 LP 和 ACC。然后只讲三个问题：