Real-time AI-curated news from 92,365+ articles across 50+ sources. Sentiment analysis, importance scoring, and key takeaways — updated every 15 minutes.
NEAR Protocol experienced an 89% price surge coinciding with a dramatic reversal in on-chain buy/sell pressure dynamics, shifting from deeply negative to +112.107 at $1.50. The rally occurred without warning signals from traditional technical indicators or derivatives markets, while NEAR's cross-chain infrastructure expanded to 30+ connected chains with institutional staking interest growing.
Singapore's Anti-Scam Centre and Cyber Investigation Branch partnered with major cryptocurrency exchanges to halt over $7 million in potential losses across two consecutive enforcement operations. The coordinated effort demonstrates growing regulatory sophistication in intercepting fraud before victims transfer funds, marking a significant shift in proactive crypto crime prevention.
Researchers introduce SMAC-Talk, a benchmark environment that extends the StarCraft Multi-Agent Challenge to evaluate how large language models coordinate and communicate in cooperative multi-agent settings. The framework tests LLM agents under realistic constraints including partial observability, decentralized control, and adversarial deception, using Qwen models to examine how reasoning, memory, and scale impact agent coordination.
Researchers propose a framework for multi-agent systems that treats disagreement as valuable information rather than error to be eliminated. The approach abstracts reasoning traces into four symbolic disagreement states and applies strategic routing rules to content moderation and AI collaboration tasks.
Researchers introduced VAMPS, a benchmark dataset of 1,168 mathematical problems designed to test whether multimodal AI models can effectively use visualization tools to solve complex algebra and calculus problems. Surprisingly, the study found that direct analytical solving consistently outperformed graph-assisted approaches across multiple models, even when visualization should theoretically help.
Researchers introduce StepPRM-RTL, a framework that enhances LLM-based RTL code generation for hardware design by combining stepwise trajectory modeling, process-reward models, and retrieval-augmented fine-tuning. The system achieves over 10% improvement in functional correctness compared to prior methods, advancing automation in hardware design workflows.
Researchers conducted mixed-methods studies on how mathematicians use AI tools to formalize proofs, finding that users prefer AI assistance while maintaining high-level control over proof discovery. A controlled user study showed participants achieved higher formalization accuracy with AI access than without, despite current tool limitations.
Researchers evaluated eight memory systems for LLM agents across five different scenarios and found that agent-controlled memory management outperforms fixed pipeline designs. The study introduces AutoMEM, a new memory harness that achieves superior cross-scenario generality by allowing agents active control over storage and retrieval operations.
Researchers introduce State-Grounded Dynamic Retrieval (SGDR), a new method enabling language agents to dynamically reuse learned skills during web automation tasks. By matching skills to both task goals and current webpage states rather than fixed skill sets, SGDR achieves 10.6% relative performance gains over existing approaches on complex multi-step web tasks.
Researchers propose a consequence-aware compute allocation system for reasoning models that prioritizes high-impact tasks based on real-world failure costs rather than just predicted difficulty. Testing on software engineering benchmarks shows the method reduces cost-weighted loss by 22-33% compared to difficulty-based routing, with a practical predictor-driven variant retaining over 90% of theoretical gains.
Trivium introduces a framework for AI agents that tracks temporal regret—how long errors persist—alongside outcome and epistemic regret to improve long-term learning. The research demonstrates that outcome-only optimization fails to correct systematic causal misunderstandings, and proposes a logarithmic-complexity intervention strategy that achieves O(log E) temporal regret across episode horizons.
AgentJet is a decoupled distributed framework for training LLM-based reinforcement learning agents across multiple nodes, enabling heterogeneous multi-agent teams and fault-tolerant execution. The system achieves 1.5-10x training speedup through context tracking optimization and automates long-horizon RL research workflows without human intervention.
Researchers introduce BioManus, an AI agent system that uses graph-based planning and standardized Model Context Protocol (MCP) servers to automate biomedical workflows. The system addresses scalability challenges by organizing bioinformatics tools into structured capability graphs rather than relying on flat prompt-based retrieval, achieving significant improvements in execution accuracy and context efficiency.
Researchers introduce MechSim, a neuro-symbolic framework that enables large language models to reason transparently about the assumptions and mechanisms underlying scientific simulators. The approach improves explainability and decision-making reliability in high-stakes simulation-driven applications by treating simulators as structured systems rather than black boxes.
Researchers developed Neetyabhas, an agent-based simulation framework that models pandemic policy decisions under real-world uncertainty, incorporating individual behavioral choices and imperfect data. Using reinforcement learning, the model demonstrates that masks and vaccines effectively reduce outbreak severity when policies account for implementation errors and measurement gaps.
Researchers have developed a machine-learning framework that learns to create admissible heuristics for optimal planning by leveraging cost partitioning and Lagrangian duality. The approach uses graph neural networks with Weisfeiler-Leman algorithms to generate cost weights that guarantee admissibility by construction, marking the first learned heuristic with formal optimality guarantees.
Researchers propose DMAIC-IAD, an LLM-based multi-agent system for industrial anomaly detection that combines structured planning with pre-trained judgment models. The system achieves 37.76% performance improvement over existing agentic baselines by standardizing heterogeneous data inputs and evaluating strategies without costly runtime execution.
Researchers propose MONIR, a normative intermediate representation framework for automated compliance reasoning using Answer Set Programming (ASP). The system combines staged operational semantics with executable ASP compilation to evaluate regulatory adherence, demonstrated through application to Chinese ADAS (Advanced Driver Assistance Systems) regulations with LLM-assisted extraction pipelines.
BiNSGPS introduces a bidirectional neuro-symbolic framework that enables dynamic feedback loops between machine learning models and symbolic solvers for geometry problem-solving. Unlike traditional unidirectional approaches, this system allows the neural component to actively incorporate feedback and correct errors, addressing fundamental limitations in AI's ability to solve complex geometric reasoning tasks.
Researchers introduce an affinity-based reinforcement learning approach tested in the board game Fog of Love, demonstrating that localized affinities enable AI agents to balance competitive and cooperative objectives simultaneously. This advancement moves virtuous AI behavior engineering from simplified toy environments to more complex multi-agent scenarios, improving agent interpretability and performance in nuanced social settings.
Researchers introduce FALSIFYBENCH, an evaluation framework that tests whether large language models can perform inductive reasoning through hypothesis-driven discovery tasks. Testing 12 LLMs reveals that reasoning models outperform instruction-tuned models, with success primarily driven by the ability to actively falsify hypotheses rather than confirm them.
Researchers introduce a tree-based mathematical framework formalizing complementarity in human-AI interactions, proving that complementarity is theoretically achievable in regression tasks but fundamentally obstructed in classification under standard loss functions. The work provides formal conditions for when AI and human predictions can outperform individual agents.
Researchers introduce BiasGRPO, a novel framework using Group Relative Policy Optimization to mitigate social bias in Large Language Models more effectively than existing methods. The approach stabilizes training in high-variance reward landscapes by normalizing rewards across sampled completions, outperforming Direct Preference Optimization and Proximal Policy Optimization while maintaining computational efficiency.
Researchers propose constraint injection, a novel verification technique that detects missing or spurious constraints in LLM-generated optimization code. VRPCoder, an 8B model fine-tuned with this method, achieves 93% accuracy on vehicle routing problems, significantly outperforming GPT and Claude models on constraint-dense combinatorial optimization tasks.
Researchers provide a rigorous mathematical framework showing how Active Inference and Expected Free Energy (EFE) minimization can be decomposed into Variational Free Energy (VFE) minimization with explicit entropy corrections. The work clarifies the theoretical foundations of EFE-based planning by identifying which corrections are necessary for different decision-making scenarios, demonstrated through grid-world experiments.
