A study using 134,376 hourly weather observations from Ioannina, Greece (2011-2026) found that hybrid CNN-GRU models outperform traditional recurrent networks in weather forecasting, achieving a Weigh…
ATSInfer, a hybrid CPU-GPU inference system for consumer devices, boosts local large language model performance by up to 1.94 times in prefill throughput and 3.29 times in decode throughput through in…
A new voice anonymization model from researchers using a frozen wav2vec2 encoder and HiFi-GAN vocoder achieves a 2.53% word error rate while preserving content and emotion, with an equal error rate of…
EmoStyle, a framework developed by researchers, won first place in the AffectiveArt Challenge 2026 by integrating emotion into AI-generated art through affective cues and style-specific LoRA adapters.…
Eastern Europe's largest banks are deploying a multimodal AI system that pretrains a foundation transformer model on a mix of transaction histories and digital signals, unifying heterogeneous data int…
A new study of MedGemma-4B-it, an open-weight medical language model, found that 38% of tests bypassed its designed safety guardrails, with compliance jumping to 53.1% when questions were framed as a …
A new framework called PISA-CAPC (Physics-Informed Structure Anchoring with Capture-Aware Prototype Calibration) achieves a target-domain mean Macro-F1 score of 0.9257 on a ten-transmitter multi-anten…
A new framework called HCRMap reduces latency in Mixture-of-Experts models by up to 43.6% in prefill and 43.0% in decode stages compared to Hydra, addressing the expert hotness skew that causes comput…
Researchers have introduced STAMP (Step-wise Attribution of Modulated Potential), a new reinforcement learning approach that addresses the reward-credit mismatch by linking actions to rewards more dir…
RouteCast, a framework for evaluating AI models with delayed or censored feedback, achieved a preliminary discrimination AUC of 0.756 in a pilot study of 21 binary-outcome cases, while a blind LLM jud…
A breakthrough in AI concept representation uses Laguerre Geometry to precisely map concepts as regions rather than points, reshaping interpretability. Researchers have developed Geometric Lens, a met…
Researchers have introduced a sparse cache approach for sequence models that compresses data by storing only novel entries, using a DP-means clustering rule for allocation. The method matches full-att…
A new framework combining convolutional neural networks, long short-term memory, and reinforcement learning improves EEG-based brain-computer interface movement decoding, achieving a 41.5% relative ga…
AgentFootprint, a new benchmark focused on AI agent storage footprint, reveals a 6.7x difference in storage spread across seven frameworks using identical models and tasks, with configurations achievi…
A new framework called Self-Verified Reasoner (SVR-R1) allows AI models to self-verify their answers by asking themselves 'Yes or No' before finalizing a response, boosting accuracy on vision-language…
A new dataset called CRiT-QA exposes flaws in large language models' multi-hop reasoning, revealing their dependence on memorized knowledge and superficial shortcuts. The dataset introduces counterfac…
A new study on distributional reinforcement learning agents found that 40-95% of risk trade-off claims made by QR-DQN, C51, and IQN models on MinAtar were refuted at 95% confidence, suggesting that pe…
Researchers have developed a method using case-based decision theory to decompose neural network decisions into weighted sums of past training cases, revealing which historical cases most influence an…
A new framework harnesses contemplative traditions such as mindfulness and compassion to enhance ethical alignment in large language models, initially targeting mental health applications. The modular…
Agentic-DPO, a new method for training language models using offline preference supervision without costly environment rollouts, improved accuracy on tau-bench from 21.7% to 41.4% for a 9B model, matc…