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

📌 MAROKO133 Hot ai: Researchers from PSU and Duke introduce “Multi-Agent Systems A

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: Double Murder Suspect Asked ChatGPT How to Hide Body in D

  Earlier this month, we got a glimpse of the harrowing conversations that Florida State University school shooting suspect Phoenix Ikner had with ChatGPT before his deadly massacre.

  Ikner asked the chatbot how to turn off the safety switch on his weapon, what ammo to use, and even where to find the most people to kill on the university’s campus — horrific queries highlighting how some are already using AI to plan and perpetrate unconscionable crimes.

  Now, prosecutors have revealed that the prime suspect in the murder of two University of South Florida doctoral students asked ChatGPT whether to hide a human body in a dumpster.

  “What happens if a human has a put [sic] in a black garbage bag and thrown in a dumpster,” 26-year-old Hisham Abugharbieh, who has since been charged with first-degree murder of Zamil Limon and Nahida Bristy, asked the chatbot, as quoted by NBC News.

  After ChatGPT told him that it sounded dangerous, Abugharbieh replied with a grim and telling question.

  “How would they find out,” he wrote.

  The evidence in the case sounds damning. One of the suspect’s roommates saw him loading boxes into a compactor dumpster. A subsequent search revealed items that once belonged to Limon, including a student ID.

  Limon’s body was later recovered by investigators inside a heavy duty trash bag — not thrown into a dumpster, but on the side of a bridge that spans Tampa Bay, showing signs of “multiple sharp force injuries,” according to an autopsy.

  Bristy’s body has yet to be identified, despite human remains being recovered over the weekend, according to NBC.

  Abugharbieh is facing serious charges, including first-degree murder, battery, false imprisonment, and storing remains in unapproved conditions, court documents show.

  Investigators have yet to detail a possible motive, but Abugharbieh’s use of ChatGPT sheds light on a worrying new trend. AI chatbot responses to some hair-raising prompts are increasingly showing up in court filings, underlining how ubiquitous the tech has become — and how some perpetrators are frequently leaving a highly incriminating paper trail as a result.

  Roughly ten months after the fatal shooting at Florida State University, a horrific school shooting in the rural mining town of Tumbler Ridge, British Columbia, similarly implicated OpenAI. The perpetrator, 18-year-old Jesse Van Rootselaar, used ChatGPT in disturbing ways before the killings. While her account was flagged last year, OpenAI never notified law enforcement, a failure to act that has resulted in a barrage of lawsuits.

  All the negative press has clearly rattled OpenAI. In a bizarre blog post this week, the company attempted to take a reassuring tone, vowing to “learn, improve and course-correct” following “mass shootings, threats against public officials, bombing attempts, and attacks on communities and individuals.”

  More on ChatGPT murders: OpenAI Hit With Barrage of Lawsuits Over Failure to Report School Shooter Before Massacre

  The post Double Murder Suspect Asked ChatGPT How to Hide Body in Dumpster appeared first on Futurism.

  🔗 Sumber: futurism.com

  ---