📌 TOPINDIATOURS Update ai: Which Agent Causes Task Failures and When?Researchers f

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: chain.zhang@jiqizhixin.com

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

📌 TOPINDIATOURS Hot ai: Video: Humanoid robot dances to viral Bollywood track at I

A humanoid robot delivered one of the most unexpected moments at a major technology festival in India, turning a standard demonstration into a live stage performance that blurred the line between engineering and entertainment.

During Techfest 2025 at the Indian Institute of Technology Bombay, a humanoid robot appeared on the main stage and performed a choreographed dance routine to the Arabic track “Fa9la.” 

The performance unfolded before students, families, researchers, and visiting technologists attending the annual event, which is widely recognised as Asia’s largest science and technology festival.

Techfest has long served as a platform for emerging research in robotics, artificial intelligence, and advanced engineering. 

This year’s humanoid performance stood out by placing those technologies in a public, creative setting rather than a controlled laboratory environment.

Engineering meets performance

Based on its structure and motion capabilities, the robot closely resembled the Unitree G1, developed by Unitree Robotics.

The G1 stands around 4.3 feet tall and weighs approximately  77 pounds. 

Engineers designed it with a highly articulated frame, enabling coordinated movement across its arms, legs, and torso.

The humanoid platform uses electric actuators and real-time motion control algorithms to maintain balance during dynamic movement. 

Integrated depth cameras and LiDAR sensors help it perceive its surroundings and adjust posture instantly. 

Unitree positions the G1 primarily as a research and education robot, but its appearance at IIT Bombay demonstrated how those technical capabilities extend into expressive, live motion.

The choreography followed the song’s rhythm closely, with smooth transitions and consistent stability, highlighting advances in humanoid locomotion and control.

Not a first appearance

The Techfest performance was not an isolated case. In recent months, humanoid robots have increasingly appeared in live stage settings. 

Unitree G1 robots recently performed synchronized choreography during a large-scale concert in China alongside singer Wang Leehom, sharing the stage with human performers under concert lighting and live music conditions.

Such appearances reflect a broader shift in how robotics companies showcase progress.

Live performances stress-test balance, timing, and adaptability in ways traditional demonstrations cannot. 

Engineers use these environments to validate systems under unpredictable conditions, including lighting changes, sound vibrations, and uneven surfaces.

As a result, stage performances have become a proving ground for humanoid robotics rather than simple publicity events.

Music behind the moment

The track “Fa9la,” also known as “Sher-e-Baloch,” features in the recent Bollywood film Dhurandhar, directed by Aditya Dhar. 

Bahrain-based rapper Flipperachi composed the song, which gained popularity for its driving rhythm and distinctive choreography.

The film stars Indian actors R Madhavan, Ranveer Singh, and Akshaye Khanna. 

At Techfest 2025, a humanoid robot unexpectedly joined that cultural moment, reinforcing how robotics now intersects with popular media.

The performance highlighted a growing reality in robotics: humanoid machines are no longer limited to factories or research labs. 

On a festival stage in India, advanced engineering stepped into the spotlight and performed.

🔗 Sumber: interestingengineering.com