<p>To bridge the gap between neural geometry and oscillatory dynamics, we analyzed gamma-band oscillations in macaque inferior temporal cortex using large-scale electrocorticography. We identified an “oscillatory manifold”—a stable, low-dimensional geometric structure embedded within oscillatory waveforms—that encodes color information. Within this manifold, trajectories for different colors were segregated via color-specific amplitude modulations of a shared oscillatory carrier. While the absolute spatial separation of these trajectories rapidly decayed following an initial stimulus-locked transient, their scale-invariant topological shape was conserved. Although stimulus shape was decodable from gamma-band power, it lacked the stable geometric structure observed for color. This contrast confirms that the oscillatory manifold is a specific representational format rather than a generic byproduct of neural activation. Consequently, gamma oscillations provide a dynamic “geometric scaffold” safeguarding perceptual fidelity against energy fluctuations. This framework unifies representational geometry with oscillatory dynamics, offering a feature-specific perspective on how the cortex stabilizes sensory information.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Gamma oscillations provide a stable geometric scaffold for color representation in primate inferior temporal cortex

  • Chengpeng Li,
  • Isao Hasegawa,
  • Hisashi Tanigawa

摘要

To bridge the gap between neural geometry and oscillatory dynamics, we analyzed gamma-band oscillations in macaque inferior temporal cortex using large-scale electrocorticography. We identified an “oscillatory manifold”—a stable, low-dimensional geometric structure embedded within oscillatory waveforms—that encodes color information. Within this manifold, trajectories for different colors were segregated via color-specific amplitude modulations of a shared oscillatory carrier. While the absolute spatial separation of these trajectories rapidly decayed following an initial stimulus-locked transient, their scale-invariant topological shape was conserved. Although stimulus shape was decodable from gamma-band power, it lacked the stable geometric structure observed for color. This contrast confirms that the oscillatory manifold is a specific representational format rather than a generic byproduct of neural activation. Consequently, gamma oscillations provide a dynamic “geometric scaffold” safeguarding perceptual fidelity against energy fluctuations. This framework unifies representational geometry with oscillatory dynamics, offering a feature-specific perspective on how the cortex stabilizes sensory information.