Topic: agents
57 stories found
Yesterday
Measuring how AI agents handle everyday wet-lab science
OpenAI developed BioBasicsBench to evaluate AI's ability to process real-world laboratory data, addressing the need for better performance in practical scientific tasks. This benchmark is crucial as it highlights AI's readiness for integration into actual research environments.
{\tau}-bench-LILT: Evaluating Conversational Agents on Realistic Multilingual Tool-Use Tasks
A new evaluation framework, ${\tau}$-bench-LILT, has been developed to assess conversational agents using realistic multilingual tasks, addressing the current English-centric focus which limits the effectiveness of these tools in non-English speaking regions. This matters because it ensures that organizations deploying such agents can better understand and improve their performance across different languages and cultures.
Perplexity launches AI agents for work
Perplexity has launched AI agents designed to assist with various tasks at work, aiming to enhance productivity and efficiency. This move is significant as it addresses growing demands for automation in the workplace, potentially transforming how professionals manage their daily activities.
Formalizing inverse problems in the age of AI: possibilities, pitfalls, and an invitation
A new Lean library called AutoIP formalizes inverse problems using proof assistants, leveraging recent advancements in AI to make mathematical proofs more accessible. This development is significant as it could bridge theoretical mathematics with practical applications, while also highlighting potential pitfalls in integrating AI into formal mathematical proofs.
Show HN: Agentic Identity Broker, OAuth Delegation for AI Agents
A new service called Agentic Identity Broker has been introduced to facilitate OAuth delegation for AI agents, allowing them to authenticate and access resources more flexibly. This development is crucial as it enhances security and usability in the growing ecosystem of AI-driven applications.
Thursday, November 5, 2026
GPT‑6 Cyber: Advancing the frontier of aligned cybersecurity agents
OpenAI has developed GPT-6 Cyber, a cybersecurity model aimed at identifying vulnerabilities and creating fixes while adhering to authorized boundaries. This advancement is crucial as it enhances the capabilities of aligned AI in protecting digital infrastructure.
Le Chonk brings the brains. Studio puts it to work.
Mistral Studio and Le Chonk enable faster development of enterprise agents, workflows, and applications by providing full control over models, data, and infrastructure. This advancement is crucial for businesses seeking to enhance their operational efficiency through advanced automation technologies.
Launch HN: Rivet (YC W23) – Scale and Orchestrate Agents on S3
Rivet, a company launched from Y Combinator's Winter 2023 cohort, aims to simplify the management of agents running on Amazon S3, enabling more efficient scaling and orchestration. This tool is crucial for organizations looking to enhance their cloud infrastructure operations by providing better control over distributed tasks and data processing.
Cognitive Bias Signatures in Search-Equipped LLM Agents: A Stage-Resolved Analysis
A study examines how initial conditions influence the performance of large language model (LLM) agents using search capabilities, finding that these conditions can introduce biases similar to those in human cognition. This research is crucial as it highlights potential limitations in LLMs, suggesting the need for more controlled and unbiased information retrieval processes.
Wednesday, November 4, 2026
YC's Garry Tan uses AI agents so much, he spends up to $200K a year on tokens
Garry Tan, a partner at Y Combinator, heavily relies on AI agents, spending as much as $200,000 annually on tokens, highlighting the growing importance and cost of AI tools in professional settings.
57 of 57 items shown. Sources: 169 days indexed.