Researchers propose a human-in-the-loop deep learning framework for color reconstruction of historical lenticular films, addressing failures of automated methods on curved lenticules, low-contrast, or badly captured regi…
Alibaba Group's Qwen team introduced Qwen-3D, a geometry-aware large multimodal model that compresses visual information using multi-view geometric cues and 3D Rotary Positional Embeddings, enabling efficient long-horizo…
Researchers propose Meta-Persona Anchoring with Filtered Temperature Scaling (FTS) to reduce the 'Artificial Hivemind' effect in large language models, cutting average pairwise cosine similarity from approximately 0.85 t…
Researchers propose Urban-Agent, a tool-augmented agent framework that combines a large language model with code execution, API calls, and Model Context Protocol to handle cross-system urban tasks. In evaluations using t…
A systematic review of 2,643 records found that only 19 of 57 full texts substantively integrate both legal and technical perspectives on Explainable AI (XAI) and the EU Right to Explanation, revealing a significant gap …
A new paper on arXiv (2608.02704v1) introduces Predictive Set Theory (PST), a formal generative framework that reconstructs cognitive architecture from first principles, anchoring cognition in a minimal set of operations…
Researchers introduced BLAZE (Bridging Literature, Agents, and Zero-gap Experimentation), a paradigm of socialized scientific intelligence that integrates persistent knowledge, collective reasoning, empirical validation,…
Researchers introduced VeriTrace, a multi-agent system that achieves 100% Pass@1 on the VerilogEval-V2 benchmark, the first system to attain perfect functional correctness. VeriTrace's Inspector agent operates over a com…
A new arXiv study (2608.02940v1) finds that a dense pruning score with 0.906 split-half reliability predicted a 16.1% gain but selected an endpoint 6.0% and 7.7% worse than two controls, modeling the gap via information …
Researchers from the CPS Research Group introduced neurosymbolic Hierarchical Reinforcement Learning with Incremental Knowledge (InK), combining symbolic planning with updatable knowledge and low-level neural modules, ac…
A new arXiv paper (2608.02996v1) argues that Latin America lacks a benchmark layer for native AI development, which prevents public institutions from independently evaluating foreign AI systems and companies from optimiz…
Singapore startups are developing technologies for fact-checking, cancer research, and rehabilitation to address global challenges, as highlighted by The Business Times in a profile of three companies at JTC Corp's Launc…
Foreign Affairs reports that 2026 may mark the arrival of the AI revolution in world politics, with U.S. officials including President Donald Trump touting American leadership while worrying about dangers such as AI cybe…
Researchers propose PI-Mem (Parallel-Iterative Memory), a mechanism that processes all chunks in parallel and iteratively refines a shared memory, enabling long-context reasoning up to 3.6 million tokens. On the HotpotQA…
A new study from arXiv (2608.03532v1) finds that social biases in large language models transform rather than transfer across languages, with stereotype rates shifting by up to 12 percentage points on specific axes when …
Researchers propose MDLMPE, the first positional encoding method for masked diffusion language models that explicitly tracks the evolving revealed/masked token configuration, improving performance over conventional encod…
Researchers propose CVPO (Curriculum-guided Value-Variance Policy Optimization), a new reinforcement learning method that enhances LLM reasoning by adapting to token-level value-variance and question difficulty. The meth…
Researchers introduced VIVID (Vietnamese Idioms for Validation and Interpretation Depth), the first benchmark for evaluating figurative language understanding in Vietnamese, comprising 1,636 idioms and proverbs annotated…
A new study from arXiv (2608.03233v1) evaluating ten state-of-the-art open- and closed-source large language models finds that LLMs exhibit a concerning tendency toward domain homogeneity in analogy making, often drawing…
A new study from arXiv (2608.03388v1) introduces the Alien Abduction game, an interactive probe revealing that large language models (LLMs) struggle with active multi-turn information acquisition for abductive inference.…