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

📌 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 Hot ai: Helium isotope could help pinpoint location of gold deposits t

  Scientists have found that with the help of a new chemical analysis, they could pinpoint the location of buried gold deposits beneath Scotland and Ireland. Sophisticated new chemical analysis of gases trapped in rocks for millions of years has cast new light on the origin of the gold deposits beneath Scotland and Ireland.

  “The presence of deep helium in all the deposits in the Caledonian orogenic belt is a clear sign that mantle melting is essential for the formation of this globally important type of gold deposit,” said Professor Fin Stuart from the University of Glasgow and SUERC led the project. 

  “Whether it explains the origin of other technology-critical metals is now an open question.”

  Mass spectrometric analysis

  The research team revealed that they used mass spectrometric analysis of samples of gold-bearing sulfide minerals from ore deposits in the Caledonian mountain belt to reach the surprising conclusion that gold originates in the deep Earth. Their finding may help resolve a long-standing debate about the origin of some the world’s major gold deposits.

  The team used helium isotopes to determine the contribution of mantle heat in driving the ore fluids responsible for major gold deposits in the Laurentian Caledonides of Britain and Ireland, including all active mines (Cononish, Curraghinalt, and Cavanacaw), many of which are tentatively classed as orogenic.

  The 3He/4He of fluids in Au-bearing sulfides (0.09−3.3 Ra) require a significant contribution from exsolved magmatic volatiles, implying that mantle heat is intrinsic to ore formation, according to a paper published in the journal Geology.

  Major gold deposits in the Caledonian belt

  The research team highlighted that the major gold deposits in the Caledonian belt are intimately related to the melting of mantle beneath the colliding crustal plates that produced the huge granite bodies in Scottish highlands. Their findings are based on high precision mass spectrometric analysis of the gases trapped in gold-rich sulfide minerals.

  The Caledonian mountain belt extends for 1,800 kilometres from the Appalachians in North America to northern Norway.  It formed 490-390 million years ago when the ancient continental plates of Laurentia, Baltica, and Avalonia collided, driven by forces deep in the Earth’s interior.   

  Scientists have argued for decades whether the mineral deposits in the world’s largest major mountain belts are a result of melting of hot rock below the Earth’s crust or whether the metals were mobilised by hot fluids released during the heating and deformation of crustal rocks during tectonic upheaval. 

  “These novel helium isotope signatures may serve as a key indicator for the identification of major mineral systems worldwide,” said Dr Calum Lyell, Exploration Geologist at Western Gold Exploration and lead author of the study.

  The team underlined that the trace amounts of helium dissolved in the ancient ore fluids is mainly from the Earth’s mantle. The analysis using sophisticated mass spectrometers at the Scottish Universities Environmental Research Centre (SUERC) shows for the first time that all deposits, irrespective of their size or age, contain helium with an isotopic composition indicating an origin in the melting of the Earth’s mantle. This, in turn, implies that the heat for driving the circulation of the hot gold-rich fluids also originated in the deep Earth.

  The team note that the proportion of deep-sourced helium and the temperature of the mineralising fluids appear to correspond to the size of the gold deposit. The study has two important conclusions. Firstly it implies that the gold ultimately originates in the mantle, not the crust. Secondly, it suggests that helium isotopes provide a simple geochemical method for determining the size of prospective gold deposits, according to a press release.

  🔗 Sumber: interestingengineering.com

  ---