MAROKO133 Hot ai: Which Agent Causes Task Failures and When?Researchers from PSU and Duke

📌 MAROKO133 Update ai: Which Agent Causes Task Failures and When?Researchers from

Share My Research is Synced’s column that welcomes scholars to share their own research breakthroughs with over 1.5M global AI enthusiasts. Beyond technological advances, Share My Research also calls for interesting stories behind the research and exciting research ideas. Contact us: [email protected]

Meet the authors
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 failure-responsible agent and the decisive error step that led to the task’s failure.
2. 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 models, like OpenAI o1 and DeepSeek R1, find this task challenging.- This h…

Konten dipersingkat otomatis.

🔗 Sumber: syncedreview.com

📌 MAROKO133 Hot ai: Advanced 3D tech with architectural expertise helps preserve N

Extremely important Buddhist sites in Nepal are under threat. Researchers have come together to document and analyze these buildings to ensure the legacy endures.

Dolpo, as an isolated, high-altitude region in northwest Nepal, preserves a special, traditional Tibetan culture with a direct lineage to the Bon religion, which predates Buddhism and persists in the region to this day. The BBC called the lower Dolpo circuit a 220km trek to the “roof of the world.”

“To shed light on the mainly undocumented Buddhist architecture of this region, we launched our first research project in 2018. As a result, the architectural documentation of eighteen Buddhist sites in Dolpo is now available for the first time, based on the on-site assessment and survey during four field missions between 2018 and 2023,” said researchers in the journal the Heritage.

The Institute of Architectural Theory, History of Art and Cultural Studies, and the Institute of Engineering Geodesy and Measurement Systems at Graz University of Technology (TU Graz) pioneered the landmark project.

Focusing on a small selection of the estimated 50-60 monasteries, with Shey Gompa being the most famous, they conducted the first in-depth investigation of these World Heritage sites.

The aim of their comprehensive architectural documentation is to gain new insights into special architectural features to promote the preservation of these buildings in the long term as earthquakes, landslides, and planned infrastructure projects such as the Chinese Belt and Road Initiative have endangered these sacred examples of Buddhism with even older affiliations.

A holistic and comparative study, authors continued, would help them answer a myriad of questions as these structures and sites remain largely unresearched, such as: how were the sites for sacred structures even chosen?

A seriously important architectural survey

During four research visits, which occurred between 2018 and 2023, the TU Graz research team studied 18 Buddhist sites in Nepal, “of which 16 assemblies have already been analyzed and surveyed.”

“They are part of a sacred landscape that has developed over centuries,” says lead researcher Carmen Auer.

“The choice of location, the type of building, and the orientation of the buildings are shaped by traditional narratives, geographical conditions, and symbolic representations.”

Collaborating with local populations and referring to prior documentation, researchers understood that “the choice of a site for a temple or monastery is always linked to a certain narrative, a founding story, a directive from the gods or from the souls of nature, who make their will clear through certain signs.”  

They also pointed out that by aligning the temple or sanctuary with the direction of the solar orbit, the believer achieves a close personal relationship with the cosmos. Most building sites in Dolpo are notably located on ascending mountain slopes.

Ensuring the preservation of architectural legacy

Throughout their groundbreaking study, TU Graz researchers continued to analyze the materials used, the names attributed to these sites, as well as their architecture, to derive the purpose behind the choices made.

Some of these sites include more than one building with notable interior detailing or decorations. Others possessed different functions, one of which was serving as a refuge for monks to retreat and train. Overall, researchers discovered a rich world in one of the most remote regions in the world — one that brims with artistic and architectural integrity.

To conclude,“ the preservation of cultural heritage is essential for the region, both for the identity of its own culture and tourism. There is little support from the government for all the challenges it faces, such as social changes, the lack of financial resources, inappropriate restorations, natural disasters, and climate change.” Researchers hope that their initiative will assist in ensuring that these unique sacred sites in Nepal will survive the turbulent times they continue to endure.

Researchers have documented, analyzed, and measured buildings as part of several expeditions and preserved some of the temple complexes as 3D computer models – and could thus also have contributed to the preservation of the real buildings, according to details available.

Read the study in Heritage.

🔗 Sumber: interestingengineering.com