MAROKO133 Hot ai: Researchers from PSU and Duke introduce “Multi-Agent Systems Automated F

📌 MAROKO133 Breaking ai: Researchers from PSU and Duke introduce “Multi-Agent Syst

Share My Research is Synced’s column that welcomes scholars to share their own research breakthroughs with over 2M global AI enthusiasts. Beyond technological advances, Share My Research also calls for interesting stories behind the research and exciting research ideas. 

Meet the author
Institutions: Penn State University, Duke University, Google DeepMind, University of Washington, Meta, Nanyang Technological University, and Oregon State University. The co-first authors are Shaokun Zhang of Penn State University and Ming Yin of Duke University.

In recent years, LLM Multi-Agent systems have garnered widespread attention for their collaborative approach to solving complex problems. However, it’s a common scenario for these systems to fail at a task despite a flurry of activity. This leaves developers with a critical question: which agent, at what point, was responsible for the failure? Sifting through vast interaction logs to pinpoint the root cause feels like finding a needle in a haystack—a time-consuming and labor-intensive effort.
 
This is a familiar frustration for developers. In increasingly complex Multi-Agent systems, failures are not only common but also incredibly difficult to diagnose due to the autonomous nature of agent collaboration and long information chains. Without a way to quickly identify the source of a failure, system iteration and optimization grind to a halt.
 
To address this challenge, researchers from Penn State University and Duke University, in collaboration with institutions including Google DeepMind, have introduced the novel research problem of “Automated Failure Attribution.” They have constructed the first benchmark dataset for this task, Who&When, and have developed and evaluated several automated attribution methods. This work not only highlights the complexity of the task but also paves a new path toward enhancing the reliability of LLM Multi-Agent systems.
The paper has been accepted as a Spotlight presentation at the top-tier machine learning conference, ICML 2025, and the code and dataset are now fully open-source.

Paper：https://arxiv.org/pdf/2505.00212
Code：https://github.com/mingyin1/Agents_Failure_Attribution
Dataset：https://huggingface.co/datasets/Kevin355/Who_and_When
 
 
Research Background and Challenges
LLM-driven Multi-Agent systems have demonstrated immense potential across many domains. However, these systems are fragile; errors by a single agent, misunderstandings between agents, or mistakes in information transmission can lead to the failure of the entire task.

Currently, when a system fails, developers are often left with manual and inefficient methods for debugging:
Manual Log Archaeology : Developers must manually review lengthy interaction logs to find the source of the problem.
Reliance on Expertise : The debugging process is highly dependent on the developer’s deep understanding of the system and the task at hand.
 
This “needle in a haystack” approach to debugging is not only inefficient but also severely hinders rapid system iteration and the improvement of system reliability. There is an urgent need for an automated, systematic method to pinpoint the cause of failures, effectively bridging the gap between “evaluation results” and “system improvement.”

Core Contributions
This paper makes several groundbreaking contributions to address the challenges above:
1. Defining a New Problem: The paper is the first to formalize “automated failure attribution” as a specific research task. This task is defined by identifying the

2. failure-responsible agent and the decisive error step that led to the task’s failure.

Constructing the First Benchmark Dataset: Who&When : This dataset includes a wide range of failure logs collected from 127 LLM Multi-Agent systems, which were either algorithmically generated or hand-crafted by experts to ensure realism and diversity. Each failure log is accompanied by fine-grained human annotations for:
Who: The agent responsible for the failure.
When: The specific interaction step where the decisive error occurred.
Why: A natural language explanation of the cause of the failure.

3. Exploring Initial “Automated Attribution” Methods : Using the Who&When dataset, the paper designs and assesses three distinct methods for automated failure attribution:
All-at-Once: This method provides the LLM with the user query and the complete failure log, asking it to identify the responsible agent and the decisive error step in a single pass. While cost-effective, it may struggle to pinpoint precise errors in long contexts.
Step-by-Step: This approach mimics manual debugging by having the LLM review the interaction log sequentially, making a judgment at each step until the error is found. It is more precise at locating the error step but incurs higher costs and risks accumulating errors.
Binary Search: A compromise between the first two methods, this strategy repeatedly divides the log in half, using the LLM to determine which segment contains the error. It then recursively searches the identified segment, offering a balance of cost and performance.
 
Experimental Results and Key Findings

Experiments were conducted in two settings: one where the LLM knows the ground truth answer to the problem the Multi-Agent system is trying to solve (With Ground Truth) and one where it does not (Without Ground Truth). The primary model used was GPT-4o, though other models were also tested. The systematic evaluation of these methods on the Who&When dataset yielded several important insights:

• A Long Way to Go: Current methods are far from perfect. Even the best-performing single method achieved an accuracy of only about 53.5% in identifying the responsible agent and a mere 14.2% in pinpointing the exact error step. Some methods performed even worse than random guessing, underscoring the difficulty of the task.
• No “All-in-One” Solution: Different methods excel at different aspects of the problem. The All-at-Once method is better at identifying “Who,” while the Step-by-Step method is more effective at determining “When.” The Binary Search method provides a middle-ground performance.

• Hybrid Approaches Show Promise but at a High Cost: The researchers found that combining different methods, such as using the All-at-Once approach to identify a potential agent and then applying the Step-by-Step method to find the error, can improve overall performance. However, this comes with a significant increase in computational cost.

• State-of-the-Art Models Struggle: Surprisingly, even the most advanced reasoning m…

  Konten dipersingkat otomatis.

  🔗 Sumber: syncedreview.com

  ---

  📌 MAROKO133 Breaking ai: US scientists are building autonomous robots that can lea

  Scientists at Argonne National Laboratory are developing AI-powered robotic assistants that could learn laboratory procedures directly from human researchers and eventually help automate complex scientific experiments.

  The project, called RoSA, short for Robot Scientific Assistant for Accelerating Experimental Workflows, aims to create robots capable of working alongside scientists in real laboratory environments while adapting to changing conditions and different types of experiments.

  Researchers say the effort is part of the U.S. Department of Energy’s Genesis Mission, a national initiative focused on using artificial intelligence, quantum computing, and supercomputers to speed up scientific discovery and double American research productivity within the next decade.

  Instead of programming every action manually, the Argonne team plans to train robots by observing scientists as they perform experiments. Researchers will wear sensors while carrying out laboratory tasks, allowing the system to capture movements, workflows, and decision-making patterns that robots can later imitate.

  “Robots with fine motor skills already exist but using them safely and effectively in real laboratories is still very challenging,” said Nicola Ferrier, senior computer scientist at Argonne in a release. “Our approach starts by learning directly from expert scientists as they do their work.”

  Robots learn experiments

  The recorded data will be used to develop AI models capable of teaching robots how scientific procedures are correctly performed. The researchers believe this learning-based approach could help robots adapt to dynamic lab conditions without requiring constant reprogramming.

  Ferrier is leading the robotics and computer vision side of the project, while computational scientist Arvind Ramanathan is contributing expertise in autonomous laboratories and AI-driven decision-making systems.

  According to the team, the project will also classify common laboratory tasks based on their complexity and precision requirements. Different robotic systems will then be matched to the most suitable jobs.

  The researchers are exploring the use of fixed-base robotic arms, humanoid robots, and hybrid robotic systems that combine mobility with stationary precision. Before deployment in real laboratories, the systems will first be tested in virtual simulation environments.

  “Our main goal is to strengthen the basic robotics and computing tools needed so that large-scale, automated robotic systems can carry out experiments faster and more reliably,” Ferrier said.

  Faster science through AI

  The project is also expected to support another DOE-backed initiative called OPAL, or Orchestrated Platform for Autonomous Laboratories, which focuses on creating networks of self-driving laboratories capable of adapting and learning independently.

  “In OPAL, dexterous robotics – which are well coordinated and nimble – are being planned for executing biological experiments,” Ramanathan said. “By integrating AI-driven decision-making with advanced robotics, we aim to create systems that can accelerate discovery across a wide range of scientific disciplines.”

  Researchers say robotic scientific assistants could eventually handle repetitive or hazardous laboratory work while improving the speed and consistency of experiments.

  The Argonne team hopes to demonstrate a fivefold increase in task efficiency within the next year as development progresses.

  “Within the next year we hope to show a fivefold improvement in how efficiently these tasks can be completed,” Ferrier said. “In the long term, we envision robot scientific assistants that can work with existing laboratory equipment, making complex experiments both safer and more efficient. RoSA is a key step toward that future.”

  🔗 Sumber: interestingengineering.com

  ---