Research Papers and Publications

Here's a collection of my academic publications and research

Last updated: 9/8/2025•Source: Google Scholar

Yashar Talebirad

View Google Scholar Profile →

Total Papers

Published research

Citations

319

Academic impact

H-Index

Research influence

Active Years

2023-2025

Research period

Citation Analytics

Visual insights into research impact and publication trends

Citations by Year

2023

3 papers318 citations

318

2024

1 paper0 citations

2025

1 paper1 citation

Citations

Papers

Citation Velocity

citations/month

Multi-agent collaboration: Harnessing the power of intelligent llm agents

2023 • 28 months active

11.11

311 total

Wisdom of the Machines: Exploring Collective Intelligence in LLM Crowds

2025 • 2 months active

0.5

1 total

Fast local community discovery relying on the strength of links

2023 • 25 months active

0.28

7 total

Towards Community Search in Uncertain Graphs

2024 • 21 months active

0 total

USIWO: A Local Community Search Algorithm for Uncertain Graphs

2023 • 23 months active

0 total

Average citations per month since publication

Showing top 5 papers by citation velocity

Research Impact Highlight

311

Citations on top paper

319

Total citations

97.5%

Impact from top paper

"Multi-agent collaboration: Harnessing the power of intelligent llm agents" (2023) represents the highest-impact research contribution.

Published Research

Most Cited

Multi-agent collaboration: Harnessing the power of intelligent llm agents

Y Talebirad, A Nadiri

2023

arXiv preprint arXiv:2306.03314

Abstract:

In this paper, we present a novel framework for enhancing the capabilities of large language models (LLMs) by leveraging the power of multi-agent systems. Our framework introduces a collaborative environment where multiple intelligent agent components, each with distinctive attributes and roles, work together to handle complex tasks more efficiently and effectively. We demonstrate the practicality and versatility of our framework through case studies in artificial general intelligence (AGI), specifically focusing on the Auto-GPT and BabyAGI models. We also examine the "Gorilla" model, which integrates external APIs into the LLM. Our framework addresses limitations and challenges such as looping issues, security risks, scalability, system evaluation, and ethical considerations. By modeling various domains such as courtroom simulations and software development scenarios, we showcase the potential applications and benefits of our proposed multi-agent system. Our framework provides an avenue for advancing the capabilities and performance of LLMs through collaboration and knowledge exchange among intelligent agents.

📊311 citations

Fast local community discovery relying on the strength of links

M Zafarmand, Y Talebirad, E Austin, C Largeron, OR Zaïane

2023

Social Network Analysis and Mining

Abstract:

Community detection methods aim to find nodes connected to each other more than other nodes in a graph. As they explore the entire network, global methods suffer from severe limitations when handling large networks due to their time and space complexity. Local community detection methods are based on an egocentric function aiming to find only the community containing a query node (or set of query nodes). However, existing local methods are often sensitive to which query node(s) is used to discover a particular community. Our proposed approach, called SIWO “Strong In, Weak Out,” is a novel community detection method, which can locally discover densely-connected communities precisely, deterministically, and quickly. Moreover, our experimental evaluation shows that the detected community is not dependent on the initial query node within a community. This method works in a one-node-expansion way …

📊7 citations

Wisdom of the Machines: Exploring Collective Intelligence in LLM Crowds

Y Talebirad, A Parsaee, V Ohal, A Nadiri, C Szepesvari, Y Mouje, ...

2025

First Workshop on Social Simulation with LLMs

Abstract:

The "wisdom of crowds" phenomenon shows that aggregating independent estimates can yield more accurate predictions than individual guesses. While crowd-sourcing is widely applied, using large language models (LLMs) for collective estimation is largely unexplored. This work investigates how to best form an LLM "crowd" for ambiguous vision-based estimation tasks. We explore two sources of diversity: response diversity, from sampling at various temperatures, and model diversity, from using different LLM architectures. We evaluate these approaches on three vision-based datasets: human height-weight pairs, small objects with known weights, and Amazon products with their prices. Our results show that aggregating deterministic (temperature 0) outputs from a diverse set of models is the most effective strategy, outperforming any single model and ensembles that rely on stochasticity from higher temperatures. We find that temperature-induced diversity introduces more noise than signal. The median aggregation of deterministic responses from multiple models outperformed 67% of individual guesses on average, a figure that rises to 75% when relevant context is provided, demonstrating that model diversity is the key to leveraging the wisdom of LLM crowds. By establishing core principles for forming an effective LLM crowd, this work provides a stepping stone for more complex, LLM-driven social simulations.

📊1 citation

Towards Community Search in Uncertain Graphs

Y Talebirad

2024

Unknown Venue

Abstract:

The representation of real-world relationships and entities through nodes and edges in a network has found wide applicability across diverse scientific fields. At the core of network analysis are the tasks of community detection and community search, which aim to identify distinct groups within a graph. While community detection partitions the graph on a global scale, community search focuses on a specific node or group of nodes to discover a cohesive subgraph in their vicinity. Traditionally, these networks were represented as deterministic graphs with clearly defined nodes and edges. However, as networks grow in scale, analyzing these networks becomes more challenging. Coupled with this, the emergence of uncertainty in data collection has necessitated a shift towards probabilistic modeling of these relationships, presenting a suite of new complexities and challenges. In response to these challenges, this thesis first focuses on enhancing the SIWO algorithm, initially designed for community mining in deterministic graphs, to make it suitable for processing very large graphs. We introduce a methodology to convert large graphs into a format that is more manageable by local community search algorithms, ensuring efficient processing without the need to store entire networks in main memory. This is complemented by the development of data structures and optimization techniques specifically designed to manage and process large-scale network data efficiently. Building upon these enhancements, we then present USIWO, a scalable and local algorithm for community search in unweighted uncertain graphs with edge uncertainty. USIWO starts …

📊0 citations

USIWO: A Local Community Search Algorithm for Uncertain Graphs

Y Talebirad, M Zafarmand, OR Zaiane, C Largeron

2023

Proceedings of the International Conference on Advances in Social Networks Analysis and Mining

Abstract:

Community detection and community search are both critical tasks in graph mining, each serving unique purposes and presenting distinct challenges. The former aims to partition the graph vertices into densely connected subsets, while the latter adopts a more ego-centric approach, focusing on a specific node or group of nodes to identify a densely-connected subgraph that contains these query nodes. However, many real-world networks are characterized by uncertainty, leading to the notion of uncertain or probabilistic graphs. The transition from deterministic graphs to uncertain graphs introduces new challenges. We present USIWO, an efficient and practical solution for community search in unweighted uncertain graphs with edge uncertainty. In addition to being accurate, the approach utilizes an efficient data structure for storing only the relevant parts of the network in main memory, eliminating the need to store …

📊0 citations

Citation data is automatically updated weekly via Semantic Scholar API.
For collaboration opportunities or questions about this research,get in touch.