Research separate from news
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.
A significant claim from a single source is published only after further confirmation.
A research study introduces LFS-FRAME, an ensemble method combining Kolmogorov-Arnold Networks (KAN) with XGBoost. It uses leakage-free stacking for multiclass classification. On test data, it achieves accuracies of 89.85 % and 81.74 % compared with individual baseline models.
The EMNGD preprint extends natural gradient descent to Riemannian manifolds for neural PDE solvers. The framework introduces parameter constraints during optimization and achieves higher accuracy and faster convergence than the comparison methods.
A theoretical study shows that Gaussian RBF kernels behave asymptotically like linear kernels at large bandwidth parameters. Kernel PCA eigenvalues and eigenprojections converge to classical PCA; the convergence rate depends on data 'eccentricity' — the ratio of maximum to mean distance between samples.
The research paper presents DAGForge, a web system for automated construction of causal DAGs in biomedicine. The tool uses LLMs to generate causal judgments based on literature, with verification against primary sources. It achieves high agreement with reference graphs while maintaining an auditable chain…
A survey (60 papers) analyzes diffusion models for filling missing regions in medical images. DDPM and latent diffusion models emerged as dominant approaches. Applications include denoising, training data augmentation, and anomaly detection, primarily in MRI and CT. The study highlights…
A study introduces Patient Sampling for pretraining autoregressive electronic health record models. The standard approach concatenates data from all patients into a single stream, introducing biases. The new method controls the distribution of signal across patients and improves performance on datasets…
Researchers used symbolic regression to discover explicit neural network weight update rules. Across 30 benchmarks, they found a rule that outperformed standard optimizers in 25 cases, reducing MSE by an average of 44.47 %. The new rules combine adaptive normalization, momentum-like…
A scientific paper describes LatentFlow, a visualization tool for understanding how graph neural networks organize chemical information in latent spaces. The system tracks changes in molecule clusters across model layers and training stages and links them to chemical properties. The authors are research teams…
The research article introduces REDE, a method for filtering noisy steps in long model reasoning chains. The method uses final-answer attention as a supervision signal and improves hallucination detection accuracy in long-reasoning models (LRM).
The scientific article proposes a precise technical definition of AI-native: a system in which AI autonomously rewrites its own implementations with an escalation detector, a verification procedure and a guaranteed fallback. It focuses on revision authority (who is allowed to change decisions), not on the capabilities of the underlying…
A survey of progress reward modeling in robotics. Terminal signals are insufficient; robots need feedback while performing tasks. The survey organizes the field from three perspectives: model interface, signal construction methods, and available data and benchmarks.
Research identifies Role Drift — a problem where modules in compound LLM systems deviate from their assigned roles. The authors propose Role Anchor, a regularizer for detecting and controlling behavior. Experiments showed that 86 % of optimization gains disappeared when the decomposer was kept in its role.
RIS-Kernel reduces self-attention complexity from O(N²) to O(N log N) without modifying weights. On Qwen2-1.5B-Instruct, 1% density achieved 75% accuracy at 32k tokens, outperforming the native baseline (71.88%). At 65k tokens, it runs on CPU without GPU, with a 14% gain over zero-context.
Pilot research explores whether the model Qwen2.5-7B-Instruct internally represents Colombian identity, socioeconomic status and stereotypes. Using natural language autoencoders, researchers analyzed layer 20 activations on 30 examples (15 Spanish-English pairs). Qualitative evidence examined without…
A research paper introduces ARI, an approach combining large language models with retrieval-augmented generation (RAG) to restore damaged historical documents. It focuses particularly on recovering named entities from external sources. Tests on Korean historical documents demonstrated…
Research identifies obstacles to training deeper Logic Gate Networks and introduces Input-Anchored LGN, which consistently improve accuracy up to 100+ layers. The solution connects every layer to the original input, stabilizing training and preserving computational paths. Tested on MNIST…
The research paper proposes a framework for detecting when multiple people move together based on heterogeneous data. It combines an LLM for extracting names and addresses, semantic embeddings, and graph reasoning. On synthetic data, recall increased by 8–15 % and the F1-score by 6–8 % compared with the pairwise baseline.
A scientific paper on arXiv proposes a methodology for deciding whether to add a learned adapter to a frozen control policy. Tests on the Go2 robot with 200 iterations show minimal actual improvement (5.2 % of potential, 0.55 % detectable gain); direct queries return a decision of “no”, the VGCC and MPC strategies…
AI Radar monitors Czech and international sources every day, looking for changes that truly deserve attention.
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.
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.
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.
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.
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.
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.
Reset history · Outage history · Subscribe in a feed reader (RSS) · Methodology and label explanations
