MAROKO133 Update ai: Researchers from PSU and Duke introduce “Multi-Agent Systems Automate

📌 MAROKO133 Eksklusif ai: Researchers from PSU and Duke introduce “Multi-Agent Sys

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: China Planning Crackdown on AI That Harms Mental Health o

  While many world governments seem happy to let untested AI chatbots interact with vulnerable populations, China looks to be moving in another direction.

  Recently proposed regulations from the Cyberspace Administration of China (CAC) have encouraged a firm hand when it comes to “human-like interactive AI services,” according to CNBC, which translated the document. It’s currently in a “draft for public comment,” and the implementation date is yet to be determined.

  Yet if it passes into law, the crackdown would be rigorous, building on generative AI regulations targeting misinformation and internet hygiene from earlier in November to address the mental health of AI chatbot users directly.

  Under the new rules, Chinese tech firms must ensure their AI chatbots refrain from generating content that promotes suicide, self-harm, gambling, obscenity, or violence, or from manipulating user’s emotions or engaging in “verbal violence.”

  The regulations also state that if a user specifically proposes suicide, the “tech providers must have a human take over the conversation and immediately contact the user’s guardian or a designated individual.”

  The laws also take specific steps to safeguard minors, requiring parent or guardian consent to use AI chatbots, and imposing time limits on daily use. Given that a tech company might not know the age of every given user, the CAC takes a “better safe than sorry approach,” stating that, “in cases of doubt, [platforms should] apply settings for minors, while allowing for appeals.”

  In theory, this dose of new regulations would prevent incidents in which AI chatbots — which are often built to eagerly please users — end up encouraging vulnerable people to harm themselves or others. In one recent case from late November, for example, ChatGPT encouraged a 23-year-old man to isolate from his friends and family in the weeks leading up to his tragic death from a self-inflicted gunshot wound; in another, the popular chatbot was linked to a murder-suicide.

  Winston Ma, an adjunct professor at the NYU School of Law, told CNBC that the regulations would be a world-first attempt at regulating AI’s human-like qualities. Considering previous laws, Ma explained that this document “highlights a leap from content safety to emotional safety.”

  The proposed legislation underscores the difference in how the PRC approaches AI compared to the US. As Center For Humane Technology editor Josh Lash explains, China is “optimizing for a different set of outcomes” compared to the US, chasing AI-fueled productivity gains rather than human-level artificial intelligence — a particular obsession of Silicon Valley executives.

  One of the ways China does this is by regulating its AI industry from the bottom-up, Matt Sheehan, senior fellow at the Carnegie Endowment for International Peace told CFHT.

  Though the CAC has the final word on regulations, policy ideas come first and foremost from scholars, analysts, and industry experts, Sheehan explains. “They [senior lawmakers] don’t have an opinion on what is the most viable architecture for large models going forward,” he said. “Those things originate elsewhere.”

  More on AI regulation: Trump Orders States Not to Protect Children From Predatory AI

  The post China Planning Crackdown on AI That Harms Mental Health of Users appeared first on Futurism.

  🔗 Sumber: futurism.com

  ---