| 2026-11-02 | dispatch | Compact AI model for astronomical spectra works in wavelength, not pixel order | braintankAI | Discovery | | |
| 2026-11-01 | dispatch | Leaner streaming model picks one voice out of crowded, noisy scenes | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Benchmark finds AI agents remember conversations but struggle to infer user habits | braintankAI | Agency | | |
| 2026-11-01 | dispatch | Robots learn faster by rewinding imagined futures inside a world model | braintankAI | Embodied | | |
| 2026-11-01 | dispatch | Hybrid model predicts car-following on snowy Norwegian highways more accurately | braintankAI | Embodied | | |
| 2026-11-01 | dispatch | Rewarding how AI searches, not just what it answers, yields sturdier search agents | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Single-cell map links rare and common DNA variants to immune gene activity | braintankAI | Morphogen | | |
| 2026-11-01 | dispatch | Self-training AI decides for itself when its practice tasks have gone stale | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Good training pairs, not scale alone, bring AI models to a shared representation | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Simulations suggest magnetic reconnection slows down in bigger systems | braintankAI | Far Field | | |
| 2026-11-01 | dispatch | Dedicated memory for AI agents often underperforms simple retrieval, study reports | braintankAI | Agency | | |
| 2026-11-01 | dispatch | Brain scans of 2,175 people find no single neural signature of emotional control | braintankAI | Interior | | |
| 2026-11-01 | dispatch | Teaching diffusion vision-language models to look at the image only when needed | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Imagined futures let robot models learn from human videos without action labels | braintankAI | Embodied | | |
| 2026-11-01 | dispatch | AI modelled on a panel of doctors beats generalist physicians in a diagnosis benchmark | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Tencent's single model parses documents, images, audio and video into one format | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Atomic defects pin electrons in place inside a two-dimensional metal | braintankAI | Substrate | | |
| 2026-11-01 | dispatch | AI story generation trails story understanding in its use of narrative theory | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Balloon-borne telescope delivers weak-lensing catalogues for 30 merging galaxy clusters | braintankAI | Far Field | | |
| 2026-11-01 | dispatch | Inherited DNA variants change how open the genome is, cell type by cell type | braintankAI | Morphogen | | |
| 2026-11-01 | dispatch | A scheduler that plans around AI agents' tool calls cuts serving delays | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Surface-shape maps help an AI video model keep a subject consistent from every side | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Causal-graph method maps how COVID-19 symptoms cluster and persist after discharge | braintankAI | Discovery | | |
| 2026-11-01 | dispatch | Robot control model anticipates motion without generating future video | braintankAI | Embodied | | |
| 2026-11-01 | dispatch | AI that judges left from right may not learn how far away things are | braintankAI | Latent | | |
| 2026-11-01 | dispatch | Human insula divides into four regions, with the front one steering the others | braintankAI | Interior | | |
| 2026-11-01 | dispatch | Tencent's Hy-Nomas generates video worlds that respond to text and camera commands | braintankAI | Embodied | | |
| 2026-11-01 | dispatch | X-rays reveal handed spin waves in manganese telluride, a candidate altermagnet | braintankAI | Substrate | | |