Franck Dernoncourt
Research Subjects
Research Interests
Franck Dernoncourt has a total of 291 co-authors. Here are the top co-authors by number of shared publications:
- Ryan Anthony Rossi
Adobe Research, San Jose, USA
24 shared publications - Trung Bui
Adobe Research
18 shared publications Thien Huu Nguyen
Department of Computer Science University of Oregon Eugene OR USA
14 shared publications- Amir Pouran Ben Veyseh
Department of Computer and Information Science University of Oregon Eugene Oregon USA
12 shared publications - Nedim Lipka
Adobe Research, San Jose, USA
10 shared publications - Doo Soon Kim
Adobe Research
9 shared publications
Latest Publications
- Top Cited PapersOpen Access
- 4 September 2026
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 58 (15) , 1-36
Towards Trustworthy Retrieval Augmented Generation for Large Language Models: A Survey
Retrieval-Augmented Generation (RAG) enhances AI-generated content by integrating external knowledge, improving relevance, and reducing hallucinations. However, RAG also introduces risks related to reliability, safety, privacy, fairness, explainability, and accountability, which impact trustworthiness. While various methods aim to address these concerns, a unified framework is lacking. This survey bridges that gap by presenting a comprehensive roadmap for trustworthy RAG systems. We provide a structured analysis of key challenges, existing solutions, and future directions across these aspects. We further organize the field around how the retrieval and generation stages and the six trustworthiness dimensions interact, demonstrating their synergies and trade-offs through a controlled cross-dimensional analysis. Additionally, we highlight downstream applications where trustworthy RAG can make a significant impact, encouraging further research and adoption in real-world AI systems.
- 4 September 2026
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 58 (15) , 1-36
Top Cited PapersOpen AccessTowards Trustworthy Retrieval Augmented Generation for Large Language Models: A Survey
Retrieval-Augmented Generation (RAG) enhances AI-generated content by integrating external knowledge, improving relevance, and reducing hallucinations. However, RAG also introduces risks related to reliability, safety, privacy, fairness, explainability, and accountability, which impact trustworthiness. While various methods aim to address these concerns, a unified framework is lacking. This survey bridges that gap by presenting a comprehensive roadmap for trustworthy RAG systems. We provide a structured analysis of key challenges, existing solutions, and future directions across these aspects. We further organize the field around how the retrieval and generation stages and the six trustworthiness dimensions interact, demonstrating their synergies and trade-offs through a controlled cross-dimensional analysis. Additionally, we highlight downstream applications where trustworthy RAG can make a significant impact, encouraging further research and adoption in real-world AI systems.
- Open Access
- 31 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-5
Personalized Auto-Research: Towards a True AI Co-Scientist
- 31 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 1-5
Open AccessPersonalized Auto-Research: Towards a True AI Co-Scientist
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 9718-9729
Benchmarking Knowledge-Extraction Attack and Defense on Retrieval-Augmented Generation (RAG)
Retrieval-Augmented Generation (RAG) has become a cornerstone of knowledge-intensive applications, including enterprise chatbots, healthcare assistants, and agentic memory systems. However, recent studies show that knowledge-extraction attacks can recover sensitive knowledge-base content through maliciously crafted queries, raising serious privacy and intellectual-property concerns. While prior work has explored individual attack and defense techniques, the research landscape remains fragmented across retrievers, generators, and evaluations based on non-standardized metrics and datasets. To address this gap, we introduce the first systematic benchmark for knowledge-extraction attacks on RAG systems. Our benchmark covers broad attack/defense strategies, representative retrieval embedding models, open/closed-source generators, (non) graph-based indexing, all evaluated under a unified framework with standardized protocols across multiple datasets spanning diverse languages. By consolidating the experimental landscape and enabling reproducible, comparable evaluation, this benchmark provides actionable insights and a practical foundation for developing privacy-preserving RAG systems in the face of emerging knowledge extraction threats.
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 9718-9729
Open AccessBenchmarking Knowledge-Extraction Attack and Defense on Retrieval-Augmented Generation (RAG)
Retrieval-Augmented Generation (RAG) has become a cornerstone of knowledge-intensive applications, including enterprise chatbots, healthcare assistants, and agentic memory systems. However, recent studies show that knowledge-extraction attacks can recover sensitive knowledge-base content through maliciously crafted queries, raising serious privacy and intellectual-property concerns. While prior work has explored individual attack and defense techniques, the research landscape remains fragmented across retrievers, generators, and evaluations based on non-standardized metrics and datasets. To address this gap, we introduce the first systematic benchmark for knowledge-extraction attacks on RAG systems. Our benchmark covers broad attack/defense strategies, representative retrieval embedding models, open/closed-source generators, (non) graph-based indexing, all evaluated under a unified framework with standardized protocols across multiple datasets spanning diverse languages. By consolidating the experimental landscape and enabling reproducible, comparable evaluation, this benchmark provides actionable insights and a practical foundation for developing privacy-preserving RAG systems in the face of emerging knowledge extraction threats.
- Open Access
- 21 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 1216-1220
Knowledge Homophily in Large Language Models
- 21 February 2026
- Published by Association for Computing Machinery (ACM)
- p. 1216-1220
Open AccessKnowledge Homophily in Large Language Models
- Open Access
Segment Length Matters: A Study of Segment Lengths on Audio Fingerprinting Performance
Audio fingerprinting provides an identifiable representation of acoustic signals, which can be later used for identification and retrieval systems. To obtain a discriminative representation, the input audio is usually segmented into shorter time intervals, allowing local acoustic features to be extracted and analyzed. Modern neural approaches typically operate on short, fixed-duration audio segments, yet the choice of segment duration is often made heuristically and rarely examined in depth. In this paper, we study how segment length affects audio fingerprinting performance. We extend an existing neural fingerprinting architecture to adopt various segment lengths and evaluate retrieval accuracy across different segment lengths and query durations. Our results show that short segment lengths (0.5-second) generally achieve better performance. Moreover, we evaluate LLM capacity in recommending the best segment length, which shows that GPT-5-mini consistently gives the best suggestions across five considerations among three studied LLMs. Our findings provide practical guidance for selecting segment duration in large-scale neural audio retrieval systems.
Open AccessSegment Length Matters: A Study of Segment Lengths on Audio Fingerprinting Performance
Audio fingerprinting provides an identifiable representation of acoustic signals, which can be later used for identification and retrieval systems. To obtain a discriminative representation, the input audio is usually segmented into shorter time intervals, allowing local acoustic features to be extracted and analyzed. Modern neural approaches typically operate on short, fixed-duration audio segments, yet the choice of segment duration is often made heuristically and rarely examined in depth. In this paper, we study how segment length affects audio fingerprinting performance. We extend an existing neural fingerprinting architecture to adopt various segment lengths and evaluate retrieval accuracy across different segment lengths and query durations. Our results show that short segment lengths (0.5-second) generally achieve better performance. Moreover, we evaluate LLM capacity in recommending the best segment length, which shows that GPT-5-mini consistently gives the best suggestions across five considerations among three studied LLMs. Our findings provide practical guidance for selecting segment duration in large-scale neural audio retrieval systems.
- Open Access
Learning to Route LLMs from Bandit Feedback: One Policy, Many Trade-offs
Efficient use of large language models (LLMs) is critical for deployment at scale: without adaptive routing, systems either overpay for strong models or risk poor performance from weaker ones. Selecting the right LLM for each query is fundamentally an online decision problem: models differ in strengths, prices fluctuate, and users value accuracy and cost differently. Yet most routers are trained offline with labels for all candidate models, an assumption that breaks in deployment, where only the outcome of the chosen model is observed. We bridge this gap with BaRP, a Bandit-feedback Routing with Preferences approach that trains under the same partial-feedback restriction as deployment, while supporting preference-tunable inference: operators can dial the performance/cost trade-off at test time without retraining. Framed as a contextual bandit over prompt features and a user preference vector, our method simulates an online feedback setting during training and adapts its routing decisions to each new prompt, rather than depending on full-information offline supervision. Comprehensive experiments show that our method consistently outperforms strong offline routers by at least 12.46% and the largest LLM by at least 2.45%, and generalizes robustly for unseen tasks.
Open AccessLearning to Route LLMs from Bandit Feedback: One Policy, Many Trade-offs
Efficient use of large language models (LLMs) is critical for deployment at scale: without adaptive routing, systems either overpay for strong models or risk poor performance from weaker ones. Selecting the right LLM for each query is fundamentally an online decision problem: models differ in strengths, prices fluctuate, and users value accuracy and cost differently. Yet most routers are trained offline with labels for all candidate models, an assumption that breaks in deployment, where only the outcome of the chosen model is observed. We bridge this gap with BaRP, a Bandit-feedback Routing with Preferences approach that trains under the same partial-feedback restriction as deployment, while supporting preference-tunable inference: operators can dial the performance/cost trade-off at test time without retraining. Framed as a contextual bandit over prompt features and a user preference vector, our method simulates an online feedback setting during training and adapts its routing decisions to each new prompt, rather than depending on full-information offline supervision. Comprehensive experiments show that our method consistently outperforms strong offline routers by at least 12.46% and the largest LLM by at least 2.45%, and generalizes robustly for unseen tasks.
- Open Access
mSCoRe: a $M$ultilingual and Scalable Benchmark for $S$kill-based $Co$mmonsense $Re$asoning
Recent advancements in reasoning-reinforced Large Language Models (LLMs) have shown remarkable capabilities in complex reasoning tasks. However, the mechanism underlying their utilization of different human reasoning skills remains poorly investigated, especially for multilingual commonsense reasoning that involves everyday knowledge across different languages and cultures. To address this gap, we propose a \textbf{M}ultilingual and Scalable Benchmark for \textbf{S}kill-based \textbf{Co}mmonsense \textbf{Re}asoning (\textbf{mSCoRe}). Our benchmark incorporates three key components that are designed to systematically evaluate LLM's reasoning capabilities, including: (1) a novel taxonomy of reasoning skills that enables fine-grained analysis of models' reasoning processes, (2) a robust data synthesis pipeline tailored specifically for commonsense reasoning evaluation, and (3) a complexity scaling framework allowing task difficulty to scale dynamically alongside future improvements in LLM abilities. Extensive experiments on eights state-of-the-art LLMs of varying sizes and training approaches demonstrate that \textbf{mSCoRe} remains significantly challenging for current models, particularly at higher complexity levels. Our results reveal the limitations of such reasoning-reinforced models when confronted with nuanced multilingual general and cultural commonsense. We further provide detailed analysis on the models' reasoning processes, suggesting future directions for improving multilingual commonsense reasoning capabilities.
Open AccessmSCoRe: a $M$ultilingual and Scalable Benchmark for $S$kill-based $Co$mmonsense $Re$asoning
Recent advancements in reasoning-reinforced Large Language Models (LLMs) have shown remarkable capabilities in complex reasoning tasks. However, the mechanism underlying their utilization of different human reasoning skills remains poorly investigated, especially for multilingual commonsense reasoning that involves everyday knowledge across different languages and cultures. To address this gap, we propose a \textbf{M}ultilingual and Scalable Benchmark for \textbf{S}kill-based \textbf{Co}mmonsense \textbf{Re}asoning (\textbf{mSCoRe}). Our benchmark incorporates three key components that are designed to systematically evaluate LLM's reasoning capabilities, including: (1) a novel taxonomy of reasoning skills that enables fine-grained analysis of models' reasoning processes, (2) a robust data synthesis pipeline tailored specifically for commonsense reasoning evaluation, and (3) a complexity scaling framework allowing task difficulty to scale dynamically alongside future improvements in LLM abilities. Extensive experiments on eights state-of-the-art LLMs of varying sizes and training approaches demonstrate that \textbf{mSCoRe} remains significantly challenging for current models, particularly at higher complexity levels. Our results reveal the limitations of such reasoning-reinforced models when confronted with nuanced multilingual general and cultural commonsense. We further provide detailed analysis on the models' reasoning processes, suggesting future directions for improving multilingual commonsense reasoning capabilities.
- Open Access
- 13 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 1360-1370
LUSIFER: Language Universal Space Integration for Enhanced Representation in Multilingual Text Embedding Models
- 13 July 2025
- Published by Association for Computing Machinery (ACM)
- p. 1360-1370
Open AccessLUSIFER: Language Universal Space Integration for Enhanced Representation in Multilingual Text Embedding Models
- Open Access
A Graph Perspective to Probe Structural Patterns of Knowledge in Large Language Models
Large language models have been extensively studied as neural knowledge bases for their knowledge access, editability, reasoning, and explainability. However, few works focus on the structural patterns of their knowledge. Motivated by this gap, we investigate these structural patterns from a graph perspective. We quantify the knowledge of LLMs at both the triplet and entity levels, and analyze how it relates to graph structural properties such as node degree. Furthermore, we uncover the knowledge homophily, where topologically close entities exhibit similar levels of knowledgeability, which further motivates us to develop graph machine learning models to estimate entity knowledge based on its local neighbors. This model further enables valuable knowledge checking by selecting triplets less known to LLMs. Empirical results show that using selected triplets for fine-tuning leads to superior performance.
Open AccessA Graph Perspective to Probe Structural Patterns of Knowledge in Large Language Models
Large language models have been extensively studied as neural knowledge bases for their knowledge access, editability, reasoning, and explainability. However, few works focus on the structural patterns of their knowledge. Motivated by this gap, we investigate these structural patterns from a graph perspective. We quantify the knowledge of LLMs at both the triplet and entity levels, and analyze how it relates to graph structural properties such as node degree. Furthermore, we uncover the knowledge homophily, where topologically close entities exhibit similar levels of knowledgeability, which further motivates us to develop graph machine learning models to estimate entity knowledge based on its local neighbors. This model further enables valuable knowledge checking by selecting triplets less known to LLMs. Empirical results show that using selected triplets for fine-tuning leads to superior performance.
- Open Access
Understanding Generative AI Capabilities in Everyday Image Editing Tasks
Generative AI (GenAI) holds significant promise for automating everyday image editing tasks, especially following the recent release of GPT-4o on March 25, 2025. However, what subjects do people most often want edited? What kinds of editing actions do they want to perform (e.g., removing or stylizing the subject)? Do people prefer precise edits with predictable outcomes or highly creative ones? By understanding the characteristics of real-world requests and the corresponding edits made by freelance photo-editing wizards, can we draw lessons for improving AI-based editors and determine which types of requests can currently be handled successfully by AI editors? In this paper, we present a unique study addressing these questions by analyzing 83k requests from the past 12 years (2013-2025) on the Reddit community, which collected 305k PSR-wizard edits. According to human ratings, approximately only 33% of requests can be fulfilled by the best AI editors (including GPT-4o, Gemini-2.0-Flash, SeedEdit). Interestingly, AI editors perform worse on low-creativity requests that require precise editing than on more open-ended tasks. They often struggle to preserve the identity of people and animals, and frequently make non-requested touch-ups. On the other side of the table, VLM judges (e.g., o1) perform differently from human judges and may prefer AI edits more than human edits. Code and qualitative examples are available at: https://psrdataset.github.io
Open AccessUnderstanding Generative AI Capabilities in Everyday Image Editing Tasks
Generative AI (GenAI) holds significant promise for automating everyday image editing tasks, especially following the recent release of GPT-4o on March 25, 2025. However, what subjects do people most often want edited? What kinds of editing actions do they want to perform (e.g., removing or stylizing the subject)? Do people prefer precise edits with predictable outcomes or highly creative ones? By understanding the characteristics of real-world requests and the corresponding edits made by freelance photo-editing wizards, can we draw lessons for improving AI-based editors and determine which types of requests can currently be handled successfully by AI editors? In this paper, we present a unique study addressing these questions by analyzing 83k requests from the past 12 years (2013-2025) on the Reddit community, which collected 305k PSR-wizard edits. According to human ratings, approximately only 33% of requests can be fulfilled by the best AI editors (including GPT-4o, Gemini-2.0-Flash, SeedEdit). Interestingly, AI editors perform worse on low-creativity requests that require precise editing than on more open-ended tasks. They often struggle to preserve the identity of people and animals, and frequently make non-requested touch-ups. On the other side of the table, VLM judges (e.g., o1) perform differently from human judges and may prefer AI edits more than human edits. Code and qualitative examples are available at: https://psrdataset.github.io