PNAS最新 | 知覺收縮偏差不只在PFC，S2早期已顯現

PsyBrain 腦心前沿 | 公眾號 PSY-Brain_Frontier

一鍵關注，點亮星標 ??

不錯過每日前沿資訊

認知神經科學前沿文獻分享

基本信息：

Title:Early emergence of perceptual biases in the secondary somatosensory cortex

發表時間：2025.12.23

Journal:Proceedings of the National Academy of Sciences（PNAS）

影響因子：9.1

獲取原文：

1. 添加小助手:PSY-Brain-Frontier即可獲取PDF版本

引言

工作記憶（working memory, WM）讓我們在刺激消失后仍能短暫保留信息，并據此做出比較與決策。經典觀點認為，前額葉皮層（prefrontal cortex, PFC）等高階區域是WM的“主舞臺”，而感覺皮層更多負責輸入。但在延遲比較任務中，行為常出現系統性偏差：對第一個刺激的估計會向整體分布的均值靠攏，即“收縮偏差”（contraction bias/central tendency）。作者此前發現，PFC群體活動在狀態空間（state space）里呈現“彎曲/壓縮”的幾何結構，對應貝葉斯觀察者模型（Bayesian observer model）里“先驗+噪聲觀測”的整合結果。

這就引出關鍵問題：這種“帶偏的表征”究竟是PFC等高階區加工后才出現，還是在更早的感覺通路就已萌芽？本研究把目光投向次級軀體感覺皮層（secondary somatosensory cortex, S2），它參與觸覺處理，也被認為與觸覺WM有關。若S2在刺激呈現甚至早期延遲就已出現類似的“幾何彎曲”，就意味著偏差可能是分布式網絡的早期產物，而非純粹由高階決策區“后期塑形”。

核心結果

1）行為學：收縮偏差直接體現在表現上。

兩只猴子都出現典型收縮偏差（contraction bias）：低f1被估高、高f1被估低，導致不同刺激類別正確率呈V形起伏（Fig. 1）。

2）模型：貝葉斯解釋成立，延遲拉長偏差更強。

貝葉斯模型能擬合各類別曲線，并顯示延遲Δ越長，f1不確定性相對更大（Σ1?Σ2變化），先驗對f1后驗的拉拽增強（Fig. 2，Table 1）。

3）神經元：S2不靠“上一試次記憶”來制造偏差。

S2單元對當前f1有信息，但對上一試次f1的信息接近基線，提示偏差不需要S2顯式保留短期歷史（Fig. 3）。

4）群體表征：S2早期就呈現“貝葉斯式幾何”。

群體軌跡距離隨f1呈S形扭曲（sigmoid），并在f1呈現期與延遲早期更貼近貝葉斯估計f1（f1,Bayes）而非真實f1；但S2隨延遲更快變噪，PFC更穩定（Fig. 5–7，重點Fig. 6）。

Fig. 1. Task design, stimulus sets, and behavior.

Fig. 2. Bayesian fits of the accuracy curves.

Fig. 3. S2 neurons exhibit significant mutual information (MI) about currenttrial f1, but not about the previous-trial f1.

Fig. 4. State-space trajectories and their relative distances.

Fig. 5. Averaged relative distances for monkey one.

Abstract

Working memory (WM) is distributed across multiple cortical areas, suggesting that behaviors relying on WM arise from interactions between these regions. In a recent study, we found that during delayed comparison tasks, the first stimulus is not represented veridically in the prefrontal cortex (PFC), but instead is encoded in a systematically warped manner–biased toward the mean of the stimulus distribution. This neural distortion, which emerges already during the stimulus presentation and persists throughout the delay period, closely mirrors a contraction bias observed in behavior. Furthermore, the behavioral responses could be explained by a Bayesian observer model, in which the observer integrates prior expectations with noisy sensory inputs. These results suggest that the geometry of PFC neural trajectories embodies Bayesian estimates that underlie biased decisions. Here, we investigate whether the secondary somatosensory cortex (S2)–a lower-level sensory area also implicated in tactile WM–exhibits a similar encoding structure. Our analyses reveal that although WM-related signals in S2 are less robust than in PFC, the neural state space in S2 shares key geometric features with that of PFC, including a similarly warped representation of stimulus values. These findings suggest that perceptual biases may originate early in the cortical processing stream and are not exclusively shaped by higherorder associative areas. More broadly, our results support a distributed organization of representational warping, in which even sensory areas contribute to the formation of bias-prone representations that guide behavior.

前沿交流|歡迎加入認知神經科學前沿交流群！

核心圖表、方法細節、統計結果與討論見原文及其拓展數據。

分享人：BQ

審核：PsyBrain 腦心前沿編輯部