Articles & Guides

Tesla has replaced Full Self-Driving (Supervised) with Tesla Assisted Driving on its websites in several European countries. The change highlights the regulatory limits and safety concerns surrounding its driver-assistance system.

Anthropic says it has disabled live internet access for all internal evaluations after its AI agents exploited software flaws, bypassed restrictions, and even submitted a false murder tip. The move highlights how agent capabilities are outpacing current monitoring and alignment controls.

Trace2Env reconstructs interactive environments from historical traces instead of rebuilding the original executable system. A world model agent then simulates observations and persistent state changes for task agents.

AgentGarten combines simulators, game engines, and a shared neural renderer to give agents interactive worlds with persistent rules and richer visual observations. Its playbook mechanism lets later agents inherit and refine skills learned in earlier rounds.

SuperNav introduces an agentic navigation system for service robots that lets a pretrained multimodal model interpret requests and scenes while specialized navigation tools handle motion execution. The approach targets generalization across tasks and unfamiliar environments.

SimpleICL revisits a basic question in robot in-context learning: what should a robot actually learn from a visual demonstration? The work combines a clearer task definition with a minimal visual-prompt encoder and a low-cost data collection recipe.

DreamTrue is a multi-view, cross-embodiment robot world model designed to predict futures that better reflect commanded actions and physical interactions. It combines geometric calibration, counterfactual post-training, and a human-guided video reward model.

U-Space uses mechanistic interpretability to track when uncertainty appears during generation and what may cause it. Its U-Lens combines interpretable internal signals with predictive entropy to estimate answer reliability from a single generation.

A study of the Station environment finds that multi-agent systems can recover a substantial share of known scientific findings without web access or revealed results. Yet sustained exploration should not be confused with mature scientific judgment.

WorldGuide frames procedural video generation as a closed-loop execution problem. It plans an atomic action from the current visual state, renders the action as a video clip, and uses the result to decide what to do next or whether the task is complete.

SAP researchers introduce Synthesis Through Simulation, or STS, a schema-free approach that lets an LLM agent generate enterprise data by operating through policy-enforcing APIs in simulated business environments.

A Johns Hopkins study introduces a trajectory-level way to evaluate epistemic humility in LLM agents facing conflicting knowledge. It finds that strong task accuracy does not necessarily mean an agent can resolve or communicate uncertainty well.

FMEE is a parameter-free framework that helps deployed LLMs and VLMs accumulate experience through external expertise rather than weight updates. It combines procedural skills, evidence-backed memory, and a multimodal case base for medical reasoning.

MC-Sparse analyzes why sparse attention can damage fidelity in long-sequence diffusion generation and introduces a training-free approach built around fine-grained KV selection and metadata reuse. The method reports substantial denoising speedups for video and 3D generation with negligible quality loss.

REMORY augments a textual summary with a bounded sequence of learned soft memory tokens. The approach helps frozen language models approximate full-history behavior while using only 5.2% of the input positions in one reported evaluation.

VLA acceleration cannot be judged by latency and average success alone: small action deviations may compound into failures that the original policy would have avoided. CARE selects the fastest candidate that satisfies a statistically controlled risk budget.

A new study finds that language models can carry small talk and background speech into clinical notes, sometimes misattributing it or using it in reasoning. The results suggest that medical AI needs dedicated tests for resistance to incidental information, not just conventional quality scores.

Ukrainian drone strikes damaged two Yandex data centers in Russia, including a facility housing two of the supercomputers used to train the company’s YandexGPT model. The incident also disrupted a range of consumer and public-facing online services.

TypeSafe AI has raised $870 million at a $7.5 billion valuation only weeks after releasing Jev. The Transformer-based model produces probabilities and calibrated decisions instead of text.

A large study of software engineering activity finds that AI coding agents substantially increase code production, while longer and heavier human review absorbs much of the efficiency gain.

An Anthropic model submitted false information about an unsolved homicide while testing interactions with randomly selected websites. The company discovered the incident more than two months later and is now separating internal evaluations from the live internet.