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

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

  ---

  📌 TOPINDIATOURS Eksklusif ai: Flax-made smart sensor fabric lets asphalt roads sel

  Scientists have developed a smart fabric embedded with sensors that can monitor the hidden health of asphalt roads in real time, in a breakthrough that could make resurfacing more sustainable, cost-effective, and less disruptive for drivers.

  To design the innovative solution, the team of researchers from the Fraunhofer Institute for Wood Research, at the Wilhelm-Klauditz-Institut (WKI) in Germany, teamed up with experts in the SenAD2 project.

  The solution is a bio-based fabric reinforced with conductive sensor wires that is embedded directly into the asphalt. Once installed, it measures strain and stress within the base layer, with AI algorithms analyzing the data to provide real-time insights into road health.

  According to the institute, these changes alter the fabric’s electrical resistance, providing a constant flow of data about the road’s condition. They hope this interaction between the sensors and AI will help assess the condition of road structures in real time.

  Smart fabric for roads

  Road maintenance has long relied on visible surface wear or destructive core drilling to assess deeper damage. While cracks and surface defects caused by traffic and environmental stress are easy to spot, detecting micro-cracks and damage in the lower layers still requires drilling and extracting core samples.

  This, in turn, can lead to costly and inefficient repairs that not only disrupt traffic flow but also shorten the lifespan of the roadway. Now, in a bid to address the issue, the researchers created a smart measurement and analysis system that can monitor the condition of the asphalt base layer nondestructively over a large area.

  “Our goal is to be able to plan over a longer period of time, to continuously monitor changes in the condition of the road and, on that basis, to establish forecasts and incorporate them into maintenance management activities,” Christina Haxter, a research scientist at Fraunhofer WKI, revealed.

  According to Haxter, the system provides continuous insights and improves the planners’ ability to decide when and where resurfacing is needed. The team’s goal is to track and predict how asphalt roads degrade. “This won’t make the roads last longer, but it will improve efforts to monitor their condition,” she elaborated.

  Roads with built-in sensors

  The novel sensor fabric is lightweight and made from flax fibers, a natural, renewable material which is inexpensive to make. It is interwoven with ultra-thin conductive wires less than a millimeter in diameter.

  The material is integrated directly into the natural fiber fabric during the weaving process, making it highly resistant to slippage or displacement. The use of thick, heavy yarns and wide spacing further stabilizes the material.

  “The fabric has to be designed in such a way that there is no breakdown of the structure in the asphalt,” Haxter explained. “The sensors must also not be damaged either during the weaving process or when the fabric is inserted into the roadbed.”

  Haxter further revealed that the fabric is produced to withstand the weight of trucks and road pavers during construction work. Made on a double rapier loom at Fraunhofer WKI, it can be manufactured in widths of 19 inches (50 centimeters) and at any desired length, making it scalable for real-world applications.

  “The fabric is designed to withstand the rigors of installation and environmental conditions, as our initial tests have shown,” Haxter concluded in a press release. Once embedded, the sensors feed their data to a roadside measurement unit, which stores and transmits the information for analysis.

  Then, AI-powered software interprets the data to identify damage patterns and estimate how the road will degrade over time. The system also includes a digital dashboard, which makes the insights accessible not just to road agencies but also to businesses, communities, and road users affected by maintenance schedules.

  After successful lab-based feasibility tests, the system is now being trialed on a flat road segment in an industrial zone. The sensor fabric spans the full width of the roadbed, with measurement nodes recording resistance changes as vehicles pass over.

  🔗 Sumber: interestingengineering.com

  ---