A preprint is a signal, not a finished product or an independently confirmed result. We therefore track research papers separately and show the actual state of supporting evidence for each one.
1381
published research events
Latest work
A significant claim from a single source is published only after further confirmation.
A June 2026 report by an MIT committee describes a decline in consultations and study groups, as well as erosion of trust between instructors and students. It recommends incorporating AI education into introductory courses, disclosing its use, and avoiding unreliable text detectors.
The study compared fine-tuning and RAG for world models in five environments. According to the authors, fine-tuning gave agents higher rewards in 15 of 20 settings, while RAG used data more efficiently. The hybrid system outperformed the other methods studied.
The authors compared the loss of LLM capabilities with pruning, quantization and distillation. According to the study, their predictor for pruning requires half as many configuration measurements and, in math and code tests, differs from a regression using all measurements by at most 0.020 nats per token.
The WakeKV method moves the state of less active attention heads to CPU memory and allows it to be restored. According to the authors, it reduces the frequency of missing entries compared with the methods evaluated; an implementation for the Mistral-7B model increased throughput while maintaining quality in LongBench tests.
Adaptive Mutual Distillation jointly trains two LLMs and continually adjusts the weights of mutual distillation according to the tasks. According to the authors, combining the two models delivered an average improvement of 2.91 points over multi-task SFT across three base models; the tests included six benchmarks.
A study of German-language RAG for the world of The Dark Eye found that enforced reasoning in German outperformed reasoning in French, but only matched the default reasoning in English. The benefit increased with richer context that accounted for document structure. The authors released the testing environment and benchmark.
The CLIMB method selects complementary pieces of textual evidence and accepts a revised answer only if estimated confidence increases. According to the authors, it improved results on Encyclopedic-VQA and InfoSeek without additional training; the available abstract does not specify the size of the improvement.
During training, Seg-OPD prioritizes a teacher model's corrections to intermediate steps. The authors report an average relative improvement in accuracy of 5.22 % over the methods compared on mathematical reasoning and competitive programming tasks.
The authors introduce Sigma models with 3 and 8 billion parameters. According to the study, they achieve results comparable to those of discrete models in mathematics and programming; continuous trajectories make it possible to control the balance between output quality and diversity and explore inference speedups.
The study introduces the MACTS-EM framework with specialized agents, dynamic task allocation and memory for transferring patterns across domains. The authors report an improvement in prediction accuracy of 8–12 %; the available abstract does not specify the methods being compared or the metrics.
The authors identified clinical concept hubs in all 11 open-weight models examined. According to the abstract, targeted manipulation of these hubs improved performance in simulated clinical settings; hub activation also predicted doctors' preferences in a blinded evaluation.
The authors introduce MintFlow, a method that modifies the intermediate state of generation without additional training so that the output meets the specified constraints. According to the abstract, it preserves the original data distribution better than the methods used for comparison; the excerpt does not provide specific numerical results.
The study introduces TasteBench, with more than 21 thousand human ratings of 215 plant-based foods and 15 thousand molecules. The best model tested achieved 66.1 % accuracy in pairwise comparisons; the median for individual raters was 65 %. Agreement among raters was low.
The authors introduced HyMLRaman, a method for identifying six drugs. The combination of EfficientNet-B3 and SVM achieved an accuracy of 96.31 % in stratified ten-fold cross-validation. Generating additional data using DDPM helped only some classifiers.
Simon Willison tested Qwen3.8 27B on a DGX Spark device on addition tasks with answers written in words. With reasoning enabled, he recorded 167 correct answers out of 169. However, he tested each combination only once; without reasoning, he ran 30 trials per combination.
Simon Willison — AI tag (leading independent LLM commentator)
Original source ↗
Research from Google Cloud AI Research addresses the problem where AI agents that optimize their own harness overfit too much to tests. The new RRSI method regulates optimization with a budget of edit changes and a filter against task-specific tricks. Result: generalization is preserved without increasing compute costs.
NASA and IBM released the Lunar Foundation Model, an open-source foundation model for lunar science. Trained on 2 million tile bundles from 17 years of data from the Lunar Reconnaissance Orbiter and other missions. The model specializes in ice and crater detection.
Part of an overview of multiple AI topics; only this event has been covered.
Aleph Alpha created a benchmark covering 967 politically sensitive topics. Tests of the Qwen, DeepSeek, and Kimi models showed that only 17–41% of responses were balanced; the rest repeat state doctrine, evade the question, or refuse to answer. The bias also shows up in unrelated questions. Nvidia Nemotron contains approx…
AI Radar monitors Czech and international sources every day, looking for changes that truly deserve attention.
MonitorsOfficial AI company blogs, specialist media, and research sources.
Selects and combinesFilters out information noise and combines articles about the same change into a single event.
Summarizes and explainsExplains significant events in English: what happened, why it matters and where the information comes from.
The result is a quick overview of what has actually changed in the AI world, rather than another stream of articles.
Use the CS/EN switch to read the same Radar in Czech or English. English content is published after its translation has been checked, so new and older items may appear later.
Everything you need to navigate the AI world
Today’s briefingAn overview of news from the past 24 hours and a main story when available. Beside them is Live · AI Flash, followed by research and links to other Radar sections. On mobile, the blocks appear one below another.
AI FlashAn ongoing feed of brief updates with an evidence status. Links lead to a Radar detail page when one is ready, otherwise to the original source. You can also find reset and outage histories here.
Practical applicationsWhat new tools and features can do, what you can try and what their actual impact could be.
Model selectionModel comparison by type of work, capabilities, price and speed.
Research and archiveA separate research overview, topic search and older events by date.
One event, everything that matters
Each row represents one event — not one article. At a glance, you can see its significance, credibility and main point.
Illustrative example, not a current news item.
Importance: ▮▮▮ majorOpenAIModels✓ 6
Agent mode is available to all paying users
Until now, the mode was available only on the highest plan; it is now available on all paid tiers without a waitlist.
▮▮▮ major · ▮▮ important · ▮ we're tracking = how significant the change is✓ 6 = six independent publishers, not the number of articles or feeds✓ official = a clear release, law or incident is substantiated by the relevant authority1 source = no independent confirmation yetbold = who is behind the changegray text = a brief summary of what happened
The detail page contains a fuller summary, its significance and original sources. Practical impact appears in the detail and the For individuals and For businesses views. An AI Flash item reaches the main selection only after it has been expanded and meets the publication rules.
The same news, two practical uses
We first summarize each event in the same way for everyone. Based on those same facts, we then explain what the change means for your own use and what it could mean for how a company operates.
For individualsWhat you can use or try, how the change can help you at work and what to watch out for.
For businessesWhat impact the change could have on processes, costs, risks and other business decisions.
Today’s briefing is the same for everyone. Pages
For individuals and For businesses
can be found in the main navigation — they select only events relevant to the given use case.