MAROKO133 Breaking ai: Researchers from PSU and Duke introduce “Multi-Agent Systems Automa

📌 MAROKO133 Update ai: Researchers from PSU and Duke introduce “Multi-Agent System

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 Eksklusif ai: Landmark US trial puts Instagram, TikTok, YouTube before

  Three of the world’s largest technology companies will face a jury in Los Angeles this week over claims that their social media platforms harm children by design.

  The trial targets Meta, ByteDance, and Google.

  Plaintiffs accuse Instagram, TikTok, and YouTube of intentionally addicting young users to boost profits.

  Jury selection begins this week in Los Angeles County Superior Court, launching what could become a defining moment for the social media industry.

  Lawyers will question dozens of potential jurors each day. The process is expected to last several days before opening arguments begin.

  The case represents the first time these companies will defend such claims in front of a jury.

  Until now, similar lawsuits stalled in pretrial motions. Legal experts say a verdict here could shape how courts handle youth-related tech lawsuits nationwide.

  Snap Inc., the parent company of Snapchat, will not take part in the trial.

  The company reached a settlement last week for an undisclosed amount.

  At the center of the trial is a 19-year-old plaintiff identified as “KGM.” Her lawsuit will act as a bellwether case. Courts use these trials to gauge how juries respond to broader legal claims.

  KGM alleges she began using social media at a young age. Over time, she says, her use became compulsive. She claims that constant engagement worsened her depression and intensified suicidal thoughts.

  Rather than focusing on specific posts, the lawsuit targets how the platforms operate. Attorneys argue that core features pushed young users to stay online longer. They say the platforms rewarded constant engagement while discouraging breaks.

  Legal scholars say this framing is intentional. By challenging product design, plaintiffs aim to avoid legal shields that often protect tech companies. Those shields include Section 230 and free speech defenses tied to user content.

  Clay Calvert, a technology policy expert at the American Enterprise Institute, said the outcome could influence thousands of similar cases waiting in courts, as reported by The Associated Press.

  Design choices questioned

  The lawsuit claims the companies borrowed tactics from gambling and tobacco industries. It argues these techniques exploited adolescent psychology.

  Court filings point to endless scrolling, algorithmic recommendations, and notification systems.

  Plaintiffs say these features create feedback loops that are difficult for young users to escape.

  According to the lawsuit, children were not accidental casualties. It claims the platforms intentionally targeted youth engagement to drive advertising revenue.

  Some experts have compared the case to the tobacco trials of the 1990s.

  Those lawsuits led to major settlements and strict limits on marketing to minors. Plaintiffs hope jurors see similar patterns here.

  Several executives are expected to testify during the trial, including Meta CEO Mark Zuckerberg.

  Companies reject claims

  The companies deny that their platforms deliberately harm children. They argue teen mental health issues are complex and multifaceted.

  Meta said recent lawsuits oversimplify serious challenges facing young people. In a public statement, the company cited academic pressure, safety concerns, economic stress, and substance abuse.

  A Meta spokesperson said the company has invested heavily in youth safety tools for years.

  Google also pushed back against the allegations. A company spokesperson said YouTube has long focused on creating safer experiences for younger users. (AP)

  The companies point to parental controls and policy updates as evidence of responsibility. Still, legal observers say jurors may focus on intent rather than safeguards.

  If plaintiffs succeed, the verdict could force sweeping changes across the social media industry.

  🔗 Sumber: interestingengineering.com

  ---