📌 TOPINDIATOURS Hot 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: 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 Update ai: Cotton fabric generates electricity from humidity; powe

Scientists have showcased that cotton can work as a power source and harvest electricity from humidity. Scientists have now demonstrated a way to transform ordinary cotton fabric into a self-sustaining electricity generator that operates day and night by drawing energy from moisture in the air.

The innovation relies on carefully engineered polymer coatings that maintain a continuous flow of ions, enabling stable electrical output without batteries or external power sources.

Fabric is coated with polypyrrole and polydopamine

The fabric is coated with polypyrrole and polydopamine, two polymers with contrasting optical and chemical properties.

Polypyrrole strongly absorbs light across a wide wavelength range and efficiently converts it into heat. When exposed to sunlight, it warms rapidly and drives fast water evaporation. Polydopamine, by contrast, reflects more light and evaporates water more slowly, allowing it to retain moisture.

Crucially, the researchers coated only half of the polypyrrole-treated fabric with polydopamine. This asymmetrical structure ensures that one side of the fabric remains relatively wet while the other continuously dries. The result is a persistent moisture gradient that drives ion transport through the cotton’s microscopic channels.

Photothermal evaporation-driven moisture generator

The team developed a photothermal evaporation-driven moisture generator (PEMG) that modifies a cotton fabric with in situ polymerized pyrrole by a polydopamine (PDA)-structured color film to form an asymmetric photothermal temperature gradient.

The team revealed that the PDA asymmetric photothermal layer not only effectively addresses the issue of insufficient sustained power generation caused by water adsorption saturation but also mitigates the unstable moisture gradient resulting from excessive water evaporation.

“White LED bulbs emit a steady stream of light for more than 24 h. Moreover, the Voc of the PEMG (six units in series) reaches 1.18 V at an average solar intensity of 1000 W m−2 and 43% RH, while the Voc is 0.72 V at night (52% RH). Meanwhile, wearable systems designed with integrated PEMG units provide sufficient power supply for electronic devices under natural conditions,” said researchers in the study.

Manufacturing process begins with untreated cotton immersed in a solution

The manufacturing process begins with untreated cotton immersed in a solution containing pyrrole monomers and other additives. Polymerization is triggered directly on the fibers, forming a conductive, black polypyrrole layer that absorbs nearly all incoming light, reported Nanowerk.

Next, half of the fabric is exposed to an alkaline dopamine solution for nearly a day. During this step, dopamine molecules self-assemble into an ultra-thin film. The thickness of this layer causes light interference effects, giving the fabric a vivid purple appearance similar to the colors seen in soap bubbles. This layer absorbs significantly less light than polypyrrole, creating a built-in thermal contrast.

Under simulated sunlight, the polypyrrole side heats up substantially more than the polydopamine side. This temperature difference accelerates evaporation on one side and continuously draws water from the other, maintaining steady ion flow.

By connecting multiple fabric units in series, researchers demonstrated continuous operation across day and night conditions. Arrays of these units powered LED lights for more than 24 hours without interruption.

Promising for wearable electronics

The technology is particularly promising for wearable electronics. When stitched into a vest, the fabric generated higher voltages during outdoor activity, as sweat supplemented ambient humidity. The energy harvested was sufficient to charge capacitors, power small lights, and even run wireless audio devices.

Mechanical testing showed that bending, friction, and washing had little effect on performance, highlighting the robustness of the polymer coatings and their compatibility with real-world use.

The researchers also explored how environmental conditions influence output. Acidic moisture increased voltage significantly, as protons act as efficient charge carriers. Certain dissolved salts further enhanced performance by increasing electron transfer at the polymer–water interface. Computational analysis confirmed that ionic solutions amplify charge exchange, leading to stronger electrical signals.

By integrating asymmetry into both thermal and chemical properties, this cotton-based system maintains its own driving force under everyday environmental conditions. Unlike batteries, it requires no recharging, contains no rigid components, and does not degrade rapidly over time.

🔗 Sumber: interestingengineering.com