Research

This article explores how AI is affecting software engineering interviews, analyzing different interview types (take-home, live exercise, presentation, actual work) across dimensions of signal quality and cost to company. It argues that AI makes take-homes too easy and live coding less relevant, recommending that companies limit AI usage in interviews to preserve signal quality, drawing parallels to classical academic evaluation models.

Richard Thackeray and Phil Snell respond to an article by Wendy Liu on using artificial intelligence, arguing that AI enhances curiosity rather than diminishing it.

Google's Preferred Sources feature is now available in AI Overviews and AI Mode, allowing you to add your favorite sites to appear more prominently in AI-powered searches, along with new carousel and 'Highly Cited' badges.

Data Formulator 0.7 is an open-source AI-powered system for enterprise data analytics that combines data connectivity, agent-guided exploration, and visualization refinement in a shared workspace.

A Vox article explores the growing movement of AI successionists who believe artificial intelligence should replace humanity as the next step in cosmic evolution, and examines the ethical and spiritual questions this raises.

Here are 12 of the biggest Google I/O 2026 keynote moments, including news about Gemini Omni, Gemini 3.5 Flash, information agents in Search, Universal Cart, Neural Expressive, Gemini Spark, and intelligent eyewear.

Google unveiled the new Coral Board at Google I/O - a compact single-board computer for on-device AI. It runs Gemma 3 270M locally and features a RISC-V based NPU.

A new analysis shows that top AI forecasters adjust their AGI timelines based on which lab is currently leading the field, with predictions swinging from earlier to later and back again as the dominant lab changes from ChatGPT to xAI/Meta/Gemini to Anthropic.

DeepSWE is a new benchmark for evaluating AI coding agents on fresh, complex software engineering tasks. It avoids data contamination, covers diverse repositories, requires significant code changes, and uses hand-written verifiers. Leading models show a wide range of performance, with GPT-5.5 achieving 70% and others lower.

CNN has filed a lawsuit against Perplexity, claiming that the startup's AI tools generate "verbatim" copies of its work, as reported earlier by CNN. The lawsuit, filed in a New York court on Thursday, also alleges that Perplexity provides users with information locked behind CNN's subscription. Perplexity, which offers an AI "answer" engine along with the AI browser Comet, is accused of ignoring CNN's efforts "to recognize or block Perplexity's unidentified crawlers" from scraping its content. "Human beings report, research, write, edit, and create the content that Perplexity takes without permission or compensation," the lawsuit claims. I … Read the full story at The Verge.

IBM and Red Hat announce Project Lightwell, a $5 billion initiative to secure open source software using AI and a team of over 20,000 engineers, establishing a trusted clearinghouse for vulnerability management.

The article argues that the key to AI-assisted software development is not better specifications or tools, but old-fashioned practices of small batches and rapid feedback loops. Data shows that faster code generation leads to bottlenecks in design, testing, and review, slowing delivery and reducing stability. The real leverage lies in reducing batch sizes and shortening feedback cycles.

Explore seven practical AI projects that automate real workflows, including job search, web research, investment research, market trend analysis, invoice processing, chart digitization, and personalized exercise training.

This article contrasts the sense of connection from the early web with the isolating experience of modern AI, arguing that while AI is a useful tool, it cannot replace human interaction, and questions whether AI has genuinely social applications.

Major AI models exhibit a secular-rational bias, ignoring religious perspectives in ethical questions. All tested models show a negative view of Jehovah's Witnesses, according to a study by a consortium of religious universities.

This article is the seventh in a series on agentic engineering and AI-driven development, focusing on context management in AI sessions. The author shares a personal experience with Gemini forgetting earlier notes, introduces the concept of context compaction, and provides four practical techniques: split discovery from documentation, use handoff documents, give acceptance criteria rather than procedures, and use spec documents as bridges. These techniques apply to both developers and regular users, helping reduce frustration caused by AI forgetting.

Hermes Desktop is a cross-platform desktop app that bundles a Python runtime, hermes-agent (a self-improving AI agent), and hermes-web-ui (a Vue 3 + Koa chat dashboard) into a single Electron application, requiring no separate Python or Node installation. It integrates with DingTalk and is powered by DeepSeek.

Perplexity AI open-sourced a Rust reimplementation of their Unigram tokenizer, achieving 5x lower latency than Hugging Face's tokenizers crate and reducing CPU utilization by 5-6x in production. The optimizations include double-array trie, bitmap packing, and huge pages.

AIluminode is a wieldable pre-retrieval cognitive-orientation instrument that helps AI tools check contextual posture before acting, using route polarity (OPEN, PROTECT, AUDIT, DEFER, BLOCK) to reduce erroneous exploration and context bleed.

Axiom Math, founded by Chinese post-00s entrepreneur Hong Letong, has had 5 out of 8 AI-generated math papers accepted in peer-reviewed journals. The company raised $2 billion in March, achieving a $16 billion valuation.

The article explores how AI is driving a paradigm shift in digital product design, moving from command-driven to intent-driven interaction, and analyzes the new challenges in product management, user experience, decision logic, release cycles, risk, and value creation.

This month's AIhub digest covers AI for Science conference, lottery ticket hypothesis interview, world models discussion, transparent and trustworthy AI research, foundation model impacts report, AIES conference reflections, Robotics Café, ACL desk rejection policy, arXiv anti-AI slop policy, and more.

A group of former researchers from Google DeepMind, Apple, OpenAI, and Meta have launched a startup called Trajectory, aiming to help companies continuously improve their AI products by training on real-world user interactions. The company has raised a $15 million seed round at a $115 million valuation, led by Conviction. Trajectory's platform enables continuous learning for AI models, updating them based on real-world failures. It currently works with AI-native companies like Clay and Harvey, and plans to expand to Fortune 500 companies.

Robinhood launches Agentic Trading, allowing customers to connect their own AI agents to automate trading and credit card purchases with safety controls and a real-time activity feed.

Last month at Beijing's half marathon, a robot named Lightning beat the human world record by nearly seven minutes. This is the latest in a series of AI milestones prompting questions about robots entering everyday life. China leads the charge with a pledge to invest over £100bn in robotics over the next 20 years.

This paper introduces Simulation-Informed Diffusion (SID), a decentralized framework using constraint-aware diffusion models (CADM) to first simulate neighbors' future trajectories and then plan own trajectories under safety constraints. SID enables a minimal communication scheme triggered only in congested scenarios and outperforms baselines, scaling to 108 robots and 160 obstacles.

Researchers propose a real-time asynchronous event-based monocular odometry for planetary rovers, using an Error-State Kalman Filter to process event camera data for robust ego-motion estimation under high dynamic range lighting and computational constraints.

This paper presents a transformer-based architecture called Trinity that jointly performs class-specific semantic segmentation and class-agnostic terrain segmentation in a unified network. It segments terrain regions based purely on visual appearance without predefined labels or robot-dependent traversability scores, enabling robot-agnostic visual terrain priors for downstream tasks. The authors extend the OAISYS simulator to create the RUGDSynth synthetic dataset and provide the EXTerra real-world dataset. Experiments demonstrate the approach's effectiveness in complex outdoor environments.

Researchers introduce Speak-to-Objective, a modular agentic pipeline that uses a conditioned LLM to translate spoken or written commands into fully differentiable objective functions for assembling microparticles in a constraint-aware inverse solver and on an experimental optofluidic platform. The approach separates what to assemble from how to actuate, learns from user feedback, and demonstrates natural-language-programmable microscale assembly using laser-induced thermoviscous flows.

Many children face challenges in emotional regulation and social interaction, limiting their participation in therapeutic programs. This study explores engagement strategies for a tactile robot supporting children with anxiety disorders, comparing synthetic emotional feedback and point rewards. A preference study with 16 school children (ages 6-8) showed preference for emotional engagement, while a behavioral study with 14 university students (ages 20-27) found point-based systems yielded higher task accuracy (p<0.05) and sustained performance. These findings highlight age-related differences and the need to validate design assumptions through observed interaction.

A new benchmark called What-If World tests video generation models' causal reasoning by presenting paired prompts that differ in one physical detail and checking if videos diverge correctly. Evaluating nine state-of-the-art models, none exceed 52% on paired scores, with open-source models around 28%, indicating significant room for improvement. Performance correlates with visual prominence rather than physics tractability.

A prospective single-center clinical validation of the Melanoscope AI mobile dermoscopy CDSS demonstrated 88.6% agreement with expert assessment on 176 patients, with no false negatives and 88.3% specificity. The study developed a quantitative interpretability method for cascade deep learning models and a three-zone patient routing algorithm, supporting reproducible and interpretable decision-making for skin cancer screening in resource-limited settings.

This work proposes representation-conditioned diffusion models that leverage learned representations from DINOv2, DINOv3, and CLIP to generate synthetic image data. On ImageNet100, this approach outperforms class-conditioned generation by +10.76 p.p. top-1 accuracy. Scaling synthetic data can even surpass real-data training by +2.0 p.p. The method also excels in data augmentation and sample filtering, offering a promising way to augment or replace real datasets in large-scale visual learning.

This paper presents a behavioral-level activity recognition method using head-mounted IMU, going beyond basic motion primitives. The authors define five behavioral categories, construct a 160K-sample dataset from Ego4D with a four-tier quality assurance framework, and propose HiT-HAR, a 703K-parameter hierarchical model that outperforms prior models on action and scenario recognition. Observability analysis reveals locomotion is reliably observable, while object transfer and task operation benefit from temporal context; scenario-dependent signal overlap remains a challenge. Results show that architectural choices exploiting temporal context and scenario structure outperform simply scaling model size.

The 10th ABAW Workshop and Competition at CVPR 2026 advances multimodal human-centered AI by introducing new challenges including emotional mimicry intensity estimation, ambivalence/hesitancy recognition, and fine-grained violence detection, alongside traditional affect estimation and recognition tasks. The competition leverages large-scale in-the-wild datasets, and the paper track covers a broad range of topics from pose estimation to fairness and robustness.

Large language models (LLMs) are increasingly used as proxies for computational social analysis, but their ability to faithfully represent human communities' 'thick descriptions' remains a critical challenge. This paper introduces CARE (Community-Aware Reaction Evaluation), a reaction-centered framework that benchmarks LLM-simulated discourse against authentic community responses to real-world news. By characterizing a fine-grained spectrum of illocutionary tones, the diagnosis reveals a persistent 'realism gap': steering LLMs with explicit community prompts fails to inherently improve simulation fidelity. Analysis further identifies divergent behavioral signatures among frontier models, suggesting current alignment strategies are insufficient for capturing the sociolinguistic dynamics of online groups.

A new framework called FLUID adapts autoregressive language models to diffusion models for efficient parallel text generation, using Strictly Causal Alignment to reuse GPT checkpoints and Elastic Horizons to dynamically adjust denoising steps. It achieves state-of-the-art performance with significantly reduced training costs.

Researchers identify a Stability-Expressivity Gap in spoken language models when using synthetic data for low-resource languages, and propose two self-alignment frameworks (DGSA and TDSC) that recover prosodic variability and outperform commercial systems like ElevenLabs and Gemini Pro, enabling zero-shot voice cloning for Lao.

BioELX is a novel two-stage framework for cross-lingual biomedical entity linking that requires no annotated training data. It enhances SapBERT with multilingual aliases from Wikidata and uses a pre-trained LLM for context-aware disambiguation. Experiments on five benchmarks show significant improvements, especially for low-resource languages like Turkish, Korean, and Thai.

RAG-Coding is an agentic method for automated ICD-10-CM coding that orchestrates four large language model (LLM) agents and grounds decisions in external knowledge sources, improving coding accuracy and clinical compliance. On the MDACE dataset, it outperforms the best LLM baseline by 8-13% micro-F1 and 2-8% macro-F1. Compared to PLM-ICD, RAG-Coding shows higher micro recall (+11%) but lower micro precision (-6%), with comparable F1 scores. Ablation studies confirm the importance of external knowledge. The authors also release MDACE-2025, updated with expert re-annotations based on 2025 guidelines, enabling finer-grained evaluation.

This paper proposes novel techniques for inter-utterance style interpolation and intra-utterance style transition in prompt-based TTS models, addressing limitations of coarse global control. Methods include direction vector interpolation and KV-cache swapping with sliding-window attention masking. Experiments show high success rates in gender conversion and smooth style transitions within utterances.

ICG is a novel framework that integrates MLLM-based prompting with personalized preference alignment to generate high-quality, contextually relevant cover images. It extracts semantic features via meta tokens, refines them with user embeddings, and injects personalized context into diffusion models. A multi-reward learning strategy combines public rewards with a personalized preference model, eliminating the need for labeled supervision. Experiments show improvements in image quality, semantic fidelity, and personalization, boosting user appeal and recommendation accuracy.

The SignGAD framework reformulates graph anomaly detection by replacing fixed pipelines with self-designed task-conditioned workflows, and introduces a guarded final refit strategy to improve reliability under limited supervision.

This paper introduces Architecture-driven Shift (ADS), a lightweight metric for selecting pre-trained models in continual learning. ADS decouples logit shift into architecture and data dependencies, requiring only few data samples to capture shift trends. Experiments across over 175 architectures show strong monotonic correlation (Spearman's r_s ≥ 0.731) between ADS and logit shift, and ADS serves as an effective proxy for expected calibration error for reliable CL model selection across three datasets and six scenarios.

This paper introduces Metric-Aware Principal Component Analysis (MAPCA), which parameterizes PCA with a positive-definite metric matrix and positions it within the geometric deep learning framework. MAPCA interprets the metric as a geometric prior, its solutions are equivariant under the orthogonal group preserving the metric, and its spectrum is invariant. A uniqueness theorem characterizes Invariant PCA (IPCA) as the unique linear data-derived metric in the MAPCA family that is equivariant under arbitrary diagonal rescaling. The paper also discusses extensions to kernel PCA, spectral graph methods, and deep MAPCA.

This survey explores how Mixture-of-Experts (MoE) effectively addresses multimodal learning challenges from three perspectives: efficient engine, representation learner, and adapter, while identifying research gaps.

This paper presents $E^3$-Agent, an executable and evolving agent for resource management of edge AIGC. It separates a fast-path router from a slow-path LLM meta-controller, learns online from execution feedback, and adapts to unknown time-varying service-time mappings. Evaluation shows 65%-73% latency reduction over static baselines and effective stutter suppression.

Research shows that diagonal state space models (S4D) outperform more complex Mamba architectures in time series classification tasks. The authors propose lightweight variants MS4 and MS4N, which achieve higher accuracy and efficiency on 59 datasets, matching deep learning models with 2x to 10x more parameters.

This paper proposes Personalized Observation Normalization (PON) for federated reinforcement learning in heterogeneous environments. Each agent locally normalizes raw state inputs using a continuously updated running mean and variance, ensuring consistent scaling without overshadowing. Sharing normalization parameters is shown ineffective. Experiments on heterogeneous MuJoCo tasks demonstrate faster training and superior performance. Accepted at IJCNN 2025.

This paper argues that within-person behavioral variability stems from dynamic latent states, not solely from observable inputs. By intervening on the state's weighting at decision time, outcomes become causally controllable. The framework integrates six lines of evidence (causal inference, predictive processing, allostasis, attentional bottleneck, chronobiology, computational psychiatry) and a 24-month observational dataset from over 200,000 users. It yields seven testable predictions and six operational requirements for state-aware systems, with implications for digital health, education, AI personalization, and personal agency.