Timeline lokal

An IIIT-H study cautions doctors against relying entirely on AI for medical diagnosis. The researchers finds discrepancies in chest X-ray regions highlighted by vision-language models and radiologists.
Source: The Hindu
https://www.thehindu.com/news/cities/Hyderabad/iiit-h-study-caution-doctors-against-relying-entirely-on-ai-for-medical-diagnosis/article71562754.ece
#MachineLearning

Chinese state funds are increasing financing for Hua Hong Grace Semiconductor and its new Wuxi fab, pouring billions of yuan to support mature-node chip production.
Source: South China Morning Post
https://www.scmp.com/tech/big-tech/article/3370297/china-state-funds-double-down-hua-hong-legacy-chip-push?utm_source=rss_feed
#Semiconductors

SK Group Chairman Chey Tae-won stated that the group may build semiconductor plants in Gwangju alongside its Yongin chip cluster. Chey noted that production capacity needs to expand faster to keep up with surging demand.
Source: The Korea Herald
https://www.koreaherald.com/article/10897855
#Semiconductors

XREAL opened pre-orders for its AURA Android XR glasses at a starting price of $1,279. The entry model includes a compute unit with 12 GB of RAM and 256 GB of storage.
Source: TechNode
https://technode.com/2026/10/09/xreal-opens-aura-preorders-at-1279-for-android-xr-glasses/
#Tech

Researchers introduce a verification-centric perspective and a scalable benchmark for LLM-assisted peer review, emphasizing error detection.
Source: arXiv
https://arxiv.org/abs/2610.11087
#MachineLearning

Harness Compilation adapts work division between a small vision-language model and its harness. A large teacher uses execution traces to revise content.
Source: arXiv
https://arxiv.org/abs/2610.11231
#MachineLearning

Researchers introduced PARCEL, a new benchmark designed to detect legal hallucinations in large language models by checking if claims are supported by cited sources.
Source: arXiv cs.CL
https://arxiv.org/abs/2610.10971
#MachineLearning

Ecosia is replacing Mistral with open-weight AI models such as Qwen, GLM and Kimi through a partnership with the German AI platform Melious.
Source: TechNode
https://technode.com/2026/10/09/ecosia-switches-from-mistral-to-open-weight-ai-models-including-qwen-glm-and-kimi/
#AI #Mistral #Qwen

Researchers propose statement normalization to transform dialogue into short, speaker-attributed statements with source references and semantic tags.
Source: arXiv cs.CL
https://arxiv.org/abs/2610.10758
#MachineLearning

Reliance Jio has outlined a proposal to deploy about 1,600 low-Earth-orbit satellites and 23 ground stations in India.
Source: MediaNama
https://www.medianama.com/2026/10/223-jio-1600-satellites-starlink-delays/
#Tech

The National Medical Commission prohibited registered medical practitioners and hospitals from using AI-generated promotional campaigns to further commercial interests where the content misleads audiences.
Source: MediaNama
https://www.medianama.com/2026/10/223-nmc-ai-generated-medical-ads-rules/
#TechPolicy

Google and Meta have asked the government to tone down provisions of the proposed Digital Competition Bill. Over 40 Indian startups have called on the government to implement preventive regulations.
Source: MediaNama
https://www.medianama.com/2026/10/223-google-meta-digital-competition-bill-diluted/
#TechPolicy #Google #Meta

RAG-Stress is a diagnostic protocol that tests how retrieval-augmented generation models rely on evidence.
Source: arXiv cs.CL
https://arxiv.org/abs/2610.11183
#MachineLearning

Researchers evaluate Large Language Models for Machine Translation quality evaluation schemes by comparing their agreement with human annotators. Results show that both humans and models are challenged.
Source: arXiv cs.CL
https://arxiv.org/abs/2610.10918
#MachineLearning

At the start of every session, LLM agents load a fixed context file. These files can degrade performance as they grow in size.
Source: arXiv
https://arxiv.org/abs/2610.11007
#MachineLearning

TokenBank provides financial infrastructure for AI services through structured contracts. It supports service-consumption rights, cash forwards, financing, and protection claims.
Source: arXiv
https://arxiv.org/abs/2610.11333
#MachineLearning

Agentic AI systems can improve by searching longer, receiving additional support, or modifying how they propose and verify outputs.
Source: arXiv
https://arxiv.org/abs/2610.10611
#MachineLearning

Plan-and-Patch introduces a diffusion language model framework for agentic planning and repair. Diffusion achieved a higher plan repair success rate than autoregressive models.
Source: arXiv
https://arxiv.org/abs/2610.10786
#MachineLearning

BeliefScope is a controlled black-box framework for separating evidence-driven revision and directional user pressure in language models.
Source: arXiv cs.CL
https://arxiv.org/abs/2610.11305
#MachineLearning

New research tests a sociolinguistic approach to measuring user-persona fidelity in simulations, using authorship-verification stylometry and lexicon-based analysis.
Source: arXiv cs.CL
https://arxiv.org/abs/2610.10988
#MachineLearning