{
  "id": 5020337,
  "title": "Neural spiketrains and population vectors entangle neural representations",
  "url": "https://urgent.news/2026/09/01/neural-spiketrains-and-population-vectors-entangle-neural",
  "topic": "ai",
  "section": "AI",
  "published": "2026-09-01T00:00:00.000Z",
  "source": {
    "name": "bioRxiv",
    "slug": "biorxiv",
    "url": "https://www.biorxiv.org/content/10.64898/2026.08.27.744867v1?rss=1"
  },
  "original_language": "en",
  "account": "Neural recordings often involve the comparison of neural spiketrains or time bins, known as population vectors. However, when multiple variables influence neural activity, these comparisons can become entangled. The primary objective of this study is to separate these individual variables or covariates that drive neural activity, thereby unveiling their underlying structure and geometric relationships. The core concept introduced in the paper is the notion of bidirectional locality in a matrix. A matrix is deemed disentangled when its rows and columns are locally interdependent, meaning they encode the same geometric relationship and only react to a specific localized segment of the matrix. To identify such matrices, the researchers propose two methods. The first technique, called coherent projections, involves identifying non-negative projections of the neural data matrix that exhibit bidirectional locality. The second approach, referred to as clumps, entails discovering dense submatrices within the neural data matrix, each symbolizing a localized region of a solitary covariate. The researchers tested these methods on two distinct neural datasets, successfully demonstrating their capability to disentangle grid cell modules and discern a movement-driven low-dimensional structure within the motor cortex.",
  "summary": "Neural recordings are usually analyzed by comparing neural spiketrains or comparing time bins (population vectors). If multiple variables drive the neural activity these comparisons will be affected by all of them. Our aim is to disentangle these different latent variables or covariates that drive neural activity and reveal their structure and geometry. The central idea of the paper is that a…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}