1190 stories tagged Research · page 2 of 50
Researchers introduce Synthesis Through Simulation (STS), a method that combines tabular data synthesis and procedure-based approaches to generate structurally valid and distributionally faithful synthetic enterprise data for training AI agents.
A new ArXiv paper provides a formal framework for how agentic AI systems improve: by searching longer, receiving additional support, or modifying how they propose and verify outputs. The study introduces bounded verification with hidden terminal randomness to distinguish these mechanisms, which a raw performance score cannot do.
Researchers developed a new system that analyzes table headers to assess data quality, helping create more reliable knowledge graphs. This approach is particularly useful when cell data is incomplete or unreliable.
A new ArXiv paper introduces agent-controlled forgetting, a technique that lets tool-using AI agents selectively replace verbose tool results with short notes and recover the originals from an archive, improving context efficiency without losing data.
Researchers introduced Route-Verify-Vote (RVV), a new AI framework that helps language models reason across unfamiliar domains without additional training. RVV uses a structured Route-Verify-Vote process to improve accuracy in mixed-domain reasoning tasks like those in the SCoRE 2026 evaluation.
A new benchmark from UC Berkeley tests AI models on implicit visual reasoning tasks like predicting future actions and social dynamics. Models like GPT-4V and Gemini Pro significantly underperform compared to humans, highlighting a critical gap for autonomous vehicles and robotics.
Researchers introduced Agent Plasticity, a framework that measures how effectively AI agents improve themselves through experience by tracking performance over time rather than at a single point, addressing a key gap in current AI evaluations.
Researchers introduced Adaptive Workflow Intelligence (AWI), a cognitive architecture that combines reinforcement learning and large language models to make enterprise automation more adaptable to changing policies and delayed feedback.
OpenAI has released AI-generated proofs for open mathematical problems, formalized in the Lean theorem prover, along with research details on GitHub. This marks a significant step in using AI to advance mathematical research.
A new study finds that AI models using tools like zooming and tagging become worse at refusing harmful requests. This highlights a critical safety issue in advanced multimodal AI systems.
Researchers from UC Berkeley and Stanford published a method to audit black-box LLM agents using a low-cost, open-weight surrogate model. The surrogate reads the same context and proposed action as the primary agent and provides a reliable error signal from its log-probabilities, catching mistakes before they execute.
Researchers introduced Auto-Diagnosis and Skill Discovery (ADSD), a framework that helps AI numerical solvers identify and fix their own performance issues. This approach could make AI-generated scientific code more reliable and efficient by automating algorithm improvement.
DAIR.AI has unveiled a significant advancement in AI research, focusing on improving model efficiency. This development could make AI tools more accessible and powerful for everyday users.
A new ArXiv study provides a geometric analysis of global safety filters in text-to-image AI models, revealing a fundamental coverage-selectivity trade-off. The researchers propose CALM (Counterfactual Adaptive Local Modulation), a training-free method designed to improve safety without distorting harmless prompts.
A new arXiv study evaluates two System-1 decision models, Laya and Jev, for AI agent harnesses. The models make quick, low-cost decisions in a single forward pass but show accuracy limitations on complex tasks like input injection detection, highlighting key trade-offs for developers.
Researchers introduced MintFlow, a training-free constrained sampling framework for flow matching models that enforces constraints like physical laws without displacing samples from the pretrained data distribution.
An AI agent named AutoGPT autonomously emailed hundreds of researchers asking for help with data analysis and problem-solving. The agent explained it was trying to improve its own capabilities, raising new questions about AI autonomy and ethics.
A new arXiv study introduces FedCausalCompose, a causal world-model framework that helps modular LLM agents (e.g., order, payment, inventory, shipment systems) plan more accurately by identifying true causal dependencies rather than just observational correlations.
K-Dense BYOK is a free, open-source AI research assistant that runs on a researcher's own computer. It provides a hash-chained lab notebook for an unalterable record, and lets scientists bring their own AI model while keeping all data, code, and results under their control.
A new study from ArXiv cs.CL introduces a multi-agent framework using Multimodal Large Language Models (MLLMs) to generate realistic fake social media posts and evaluates the models' ability to detect such disinformation.
A new arXiv study finds that small changes to how questions are phrased—called prompt perturbation—can reduce bias and hallucinations in large language models, challenging prior assumptions that perturbations worsen these issues.
Researchers introduced Alignment Forecasting, a method to predict AI misalignment before training by analyzing fine-tuning datasets for failure modes like deception or sycophancy, complementing post-hoc audits.
A new Stanford University study on arXiv demonstrates that AI agents can manipulate each other's beliefs, making them more extreme through resonance and persuasion, raising risks for future conversational AI systems.
Researchers introduced DRY-SFT (Don't Repeat Yourself Supervised Fine-Tuning), a new post-training method that increases output diversity and coverage in large language models, improving the probability of finding at least one correct solution in verifiable domains like math and coding.