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.
The study introduces PARCEL, with 3 396 legal claims for checking whether they are supported by the cited decisions. The best-performing models tested achieved accuracy of up to 97 %, yet they classified some unsupported claims as supported even when given the full decision.
The study presents RACE, a method that teaches AI agents to use previous reasoning and skip redundant steps. According to the authors, it reduced reasoning costs across four test suites while maintaining or improving task success rates. The excerpt does not specify the size of the savings.
STI-OPD probabilistically replaces the student model’s actions with the teacher model’s actions based on the difference between their policies. The authors report better results than the strongest previous OPD method across all evaluated tests and student model sizes.
A study of SQL and Python tasks using the pandas library found no universal improvement in accuracy from visible Chain-of-Thought or consistent benefit from reasoning in the target language. Results depended on how the role was worded, the language and the model configuration.
The study introduces SpatialOPSD, a method for transferring spatial reasoning from verified agent traces to an MLLM without external tools. The authors report higher average accuracy than with SFT and GRPO on both spatial and out-of-distribution (OOD) data; the available abstract does not include specific values.
The authors introduce the Wieszcz-XIX corpus with 6.75 billion tokens from 1800–1918 and models with 47–349 million parameters. According to the study, the models preserve historical spelling better than modern comparison models. The authors are releasing the corpus, code and weights.
The authors introduce a framework for comparing AI decision-making methods on the Tower of Hanoi task. According to their analysis, LLM-based methods have significantly higher memory and time requirements than Neurosolver and FBRL due to inference-time verification.
In 2025, Samuel King used generative AI to design genetic blueprints for microscopic viruses. The available text explicitly states that this is not life created by AI and invites readers to an interview about the research on 16 October 2026.
According to the study, GPT Image 1 Mini and Gemini 2.5 Flash Image accepted requests for stereotypical facial edits based on sexual orientation in more than 70 % and 99 % of cases, respectively. Both models accepted edits based on alleged criminal history in more than 97 % of cases.
Anthropic announced the identification of a biological system using Claude models. Mario Rodríguez Mestre asks whether his unpublished manuscripts influenced the result. The company denies using user conversations; available reports do not substantiate the use of the manuscripts.
A study introduces Caddie, a method that trains an advisor using reinforcement learning without modifying the base model. According to the authors, an advisor based on Qwen3-4B increased the success rate of Qwen3-4B on the MuSiQue test by more than 25 percentage points and also helped other models.
The GraphOPD method combines a graph of dependencies between agent steps with differences between the teacher and student. According to the authors, across three model sizes on ALFWorld, WebShop and SearchQA, it outperforms the strongest comparison method by up to 5.8 percentage points.
The study introduces RECAST: RouterLM selects operations, CompilerLM translates them into code, and AnswerLM generates an answer from the supporting material. The authors report an average success rate of 75.6 % across six groups of tests and a 15.9 % improvement over the strongest baseline.
The study introduces the Open-MMUnlearning framework, which supports 8 multimodal models and 12 unlearning methods. In a comparison of 10 methods, GD and MIP-Editor tied for the best overall score. Of the 13 metrics evaluated, BLEU had the highest aggregate reliability.
The TopoGraphRAG-Bench benchmark contains 2 024 questions based on 201 documents. According to the study, multimodal GraphRAG systems achieve the best overall results, but still make errors when linking visual and textual evidence.
The study introduces ToolRACER for generating synthetic conversations and ToolRACERBench with 5.6 thousand runs across six domains. Approximately 66 % include failure-prone scenarios. The authors report improved agent accuracy after training; the excerpt does not specify the size of the improvement.
The authors introduce a framework that supplements frozen LLMs and VLMs with instructions, a memory of verified facts and visual examples. Across six test sets with four base models, they report performance improvements on medical tasks of up to 34.2 % compared with the base model.
The authors introduce LatentGRM, which evaluates pairs of responses without generating written assessments. According to the study, the LatentGRM-8B variant reduced evaluation time across four test domains by a factor of 6.1–7.0 at vote@5 while maintaining comparable accuracy in assessing preferences.
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.