arXiv subjects

Laurent Mombaerts

Publications and source records attributed to Laurent Mombaerts.

7 recordsLinked to original sources

arXiv2025

Tournament of Prompts: Evolving LLM Instructions Through Structured Debates and Elo Ratings

Prompt engineering represents a critical bottleneck to harness the full potential of Large Language Models (LLMs) for solving complex tasks, as it requires specialized expertise, significant trial-and-error, and manual intervention. This challenge is particularly pronounced for tasks involving subjective quality assessment, where defining explicit optimization objectives becomes fundamentally problematic. Existing automated prompt optimization methods falter in these scenarios, as they typically require well-defined task-specific numerical fitness functions or rely on generic templates that cannot capture the nuanced requirements of complex use cases. We introduce DEEVO (DEbate-driven EVOlutionary prompt optimization), a novel framework that guides prompt evolution through a debate-driven evaluation with an Elo-based selection. Contrary to prior work, DEEVOs approach enables exploration of the discrete prompt space while preserving semantic coherence through intelligent crossover and strategic mutation operations that incorporate debate-based feedback, combining elements from both successful and unsuccessful prompts based on identified strengths rather than arbitrary splicing. Using Elo ratings as a fitness proxy, DEEVO simultaneously drives improvement and preserves valuable diversity in the prompt population. Experimental results demonstrate that DEEVO significantly outperforms both manual prompt engineering and alternative state-of-the-art optimization approaches on open-ended tasks and close-ended tasks despite using no ground truth feedback. By connecting LLMs reasoning capabilities with adaptive optimization, DEEVO represents a significant advancement in prompt optimization research by eliminating the need of predetermined metrics to continuously improve AI systems.

cs.AI↗

arXiv2025

Topological analysis of brain dynamical signals indicates signatures of seizure susceptibility

Epilepsy is known to drastically alter brain dynamics during seizures (ictal periods), but its effects on background (non-ictal) brain dynamics remain poorly understood. To investigate this, we analyzed an in-house dataset of brain activity recordings from epileptic zebrafish, focusing on two controlled genetic conditions across two fishlines. After using machine learning to segment and label recordings, we applied time-delay embedding and Persistent Homology -- a noise-robust method from Topological Data Analysis (TDA) -- to uncover topological patterns in brain activity. We find that ictal and non-ictal periods can be distinguished based on the topology of their dynamics, independent of genetic condition or fishline, which validates our approach. Remarkably, within a single wild-type fishline, we identified topological differences in non-ictal periods between seizure-prone and seizure-free individuals. These findings suggest that epilepsy leaves detectable topological signatures in brain dynamics even outside of ictal periods. Overall, this study demonstrates the utility of TDA as a quantitative framework to screen for topological markers of epileptic susceptibility, with potential applications across species.

physics.soc-ph↗

arXiv2025

Autoencoder-based General Purpose Representation Learning for Customer Embedding

Recent advances in representation learning have successfully leveraged the underlying domain-specific structure of data across various fields. However, representing diverse and complex entities stored in tabular format within a latent space remains challenging. In this paper, we introduce DEEPCAE, a novel method for calculating the regularization term for multi-layer contractive autoencoders (CAEs). Additionally, we formalize a general-purpose entity embedding framework and use it to empirically show that DEEPCAE outperforms all other tested autoencoder variants in both reconstruction performance and downstream prediction performance. Notably, when compared to a stacked CAE across 13 datasets, DEEPCAE achieves a 34% improvement in reconstruction error.

cs.LG↗

arXiv2024

Meta Knowledge for Retrieval Augmented Large Language Models

Retrieval Augmented Generation (RAG) is a technique used to augment Large Language Models (LLMs) with contextually relevant, time-critical, or domain-specific information without altering the underlying model parameters. However, constructing RAG systems that can effectively synthesize information from large and diverse set of documents remains a significant challenge. We introduce a novel data-centric RAG workflow for LLMs, transforming the traditional retrieve-then-read system into a more advanced prepare-then-rewrite-then-retrieve-then-read framework, to achieve higher domain expert-level understanding of the knowledge base. Our methodology relies on generating metadata and synthetic Questions and Answers (QA) for each document, as well as introducing the new concept of Meta Knowledge Summary (MK Summary) for metadata-based clusters of documents. The proposed innovations enable personalized user-query augmentation and in-depth information retrieval across the knowledge base. Our research makes two significant contributions: using LLMs as evaluators and employing new comparative performance metrics, we demonstrate that (1) using augmented queries with synthetic question matching significantly outperforms traditional RAG pipelines that rely on document chunking (p < 0.01), and (2) meta knowledge-augmented queries additionally significantly improve retrieval precision and recall, as well as the final answers breadth, depth, relevancy, and specificity. Our methodology is cost-effective, costing less than $20 per 2000 research papers using Claude 3 Haiku, and can be adapted with any fine-tuning of either the language or embedding models to further enhance the performance of end-to-end RAG pipelines.

cs.IR↗

arXiv2024

VERA: Validation and Evaluation of Retrieval-Augmented Systems

The increasing use of Retrieval-Augmented Generation (RAG) systems in various applications necessitates stringent protocols to ensure RAG systems accuracy, safety, and alignment with user intentions. In this paper, we introduce VERA (Validation and Evaluation of Retrieval-Augmented Systems), a framework designed to enhance the transparency and reliability of outputs from large language models (LLMs) that utilize retrieved information. VERA improves the way we evaluate RAG systems in two important ways: (1) it introduces a cross-encoder based mechanism that encompasses a set of multidimensional metrics into a single comprehensive ranking score, addressing the challenge of prioritizing individual metrics, and (2) it employs Bootstrap statistics on LLM-based metrics across the document repository to establish confidence bounds, ensuring the repositorys topical coverage and improving the overall reliability of retrieval systems. Through several use cases, we demonstrate how VERA can strengthen decision-making processes and trust in AI applications. Our findings not only contribute to the theoretical understanding of LLM-based RAG evaluation metric but also promote the practical implementation of responsible AI systems, marking a significant advancement in the development of reliable and transparent generative AI technologies.

cs.IR↗

arXiv2020

Continuous time Gaussian process dynamical models in gene regulatory network inference

One of the focus areas of modern scientific research is to reveal mysteries related to genes and their interactions. The dynamic interactions between genes can be encoded into a gene regulatory network (GRN), which can be used to gain understanding on the genetic mechanisms behind observable phenotypes. GRN inference from time series data has recently been a focus area of systems biology. Due to low sampling frequency of the data, this is a notoriously difficult problem. We tackle the challenge by introducing the so-called continuous-time Gaussian process dynamical model, based on Gaussian process framework that has gained popularity in nonlinear regression problems arising in machine learning. The model dynamics are governed by a stochastic differential equation, where the dynamics function is modelled as a Gaussian process. We prove the existence and uniqueness of solutions of the stochastic differential equation. We derive the probability distribution for the Euler discretised trajectories and establish the convergence of the discretisation. We develop a GRN inference method called BINGO, based on the developed framework. BINGO is based on MCMC sampling of trajectories of the GPDM and estimating the hyperparameters of the covariance function of the Gaussian process. Using benchmark data examples, we show that BINGO is superior in dealing with poor time resolution and it is computationally feasible.

math.OC↗

arXiv2019

A multifactorial evaluation framework for gene regulatory network reconstruction

In the past years, many computational methods have been developed to infer the structure of gene regulatory networks from time-series data. However, the applicability and accuracy presumptions of such algorithms remain unclear due to experimental heterogeneity. This paper assesses the performance of recent and successful network inference strategies under a novel, multifactorial evaluation framework in order to highlight pragmatic tradeoffs in experimental design. The effects of data quantity and systems perturbations are addressed, thereby formulating guidelines for efficient resource management. Realistic data were generated from six widely used benchmark models of rhythmic and non-rhythmic gene regulatory systems with random perturbations mimicking the effect of gene knock-out or chemical treatments. Then, time-series data of increasing lengths were provided to five state-of-the-art network inference algorithms representing distinctive mathematical paradigms. The performances of such network reconstruction methodologies are uncovered under various experimental conditions. We report that the algorithms do not benefit equally from data increments. Furthermore, for rhythmic systems, it is more profitable for network inference strategies to be run on long time-series rather than short time-series with multiple perturbations. By contrast, for the non-rhythmic systems, increasing the number of perturbation experiments yielded better results than increasing the sampling frequency. We expect that future benchmark and algorithm design would integrate such multifactorial considerations to promote their widespread and conscientious usage.

q-bio.MN↗