Ahmed Aly

Patrick Huber, Arash Einolghozati, Rylan Conway et al. · meta-ai

Distilling conversational skills into Small Language Models (SLMs) with approximately 1 billion parameters presents significant challenges. Firstly, SLMs have limited capacity in their model parameters to learn extensive knowledge compared to larger models. Secondly, high-quality conversational datasets are often scarce, small, and domain-specific. Addressing these challenges, we introduce a novel data distillation framework named CoDi (short for Conversational Distillation, pronounced "Cody"), allowing us to synthesize large-scale, assistant-style datasets in a steerable and diverse manner. Specifically, while our framework is task agnostic at its core, we explore and evaluate the potential of CoDi on the task of conversational grounded reasoning for question answering. This is a typical on-device scenario for specialist SLMs, allowing for open-domain model responses, without requiring the model to "memorize" world knowledge in its limited weights. Our evaluations show that SLMs trained with CoDi-synthesized data achieve performance comparable to models trained on human-annotated data in standard metrics. Additionally, when using our framework to generate larger datasets from web data, our models surpass larger, instruction-tuned models in zero-shot conversational grounded reasoning tasks.

Jiaqi Wang, Xiao Yang, Kai Sun et al.

Wearable devices such as smart glasses are transforming the way people interact with their surroundings, enabling users to seek information regarding entities in their view. Multi-Modal Retrieval-Augmented Generation (MM-RAG) plays a key role in supporting such questions, yet there is still no comprehensive benchmark for this task, especially regarding wearables scenarios. To fill this gap, we present CRAG-MM -- a Comprehensive RAG benchmark for Multi-modal Multi-turn conversations. CRAG-MM contains a diverse set of 6.5K (image, question, answer) triplets and 2K visual-based multi-turn conversations across 13 domains, including 6.2K egocentric images designed to mimic captures from wearable devices. We carefully constructed the questions to reflect real-world scenarios and challenges, including five types of image-quality issues, six question types, varying entity popularity, differing information dynamism, and different conversation turns. We design three tasks: single-source augmentation, multi-source augmentation, and multi-turn conversations -- each paired with an associated retrieval corpus and APIs for both image-KG retrieval and webpage retrieval. Our evaluation shows that straightforward RAG approaches achieve only 32% and 43% truthfulness on CRAG-MM single- and multi-turn QA, respectively, whereas state-of-the-art industry solutions have similar quality (32%/45%), underscoring ample room for improvement. The benchmark has hosted KDD Cup 2025, attracting about 1K participants and 5K submissions, with winning solutions improving baseline performance by 28%, highlighting its early impact on advancing the field.

Patrick Huber, Ernie Chang, Wei Wen et al.

Efficient on-device language models around 1 billion parameters are essential for powering low-latency AI applications on mobile and wearable devices. However, achieving strong performance in this model class, while supporting long context windows and practical deployment remains a significant challenge. We introduce MobileLLM-Pro, a 1-billion-parameter language model optimized for on-device deployment. MobileLLM-Pro achieves state-of-the-art results across 11 standard benchmarks, significantly outperforming both Gemma 3-1B and Llama 3.2-1B, while supporting context windows of up to 128,000 tokens and showing only minor performance regressions at 4-bit quantization. These improvements are enabled by four core innovations: (1) implicit positional distillation, a novel technique that effectively instills long-context capabilities through knowledge distillation; (2) a specialist model merging framework that fuses multiple domain experts into a compact model without parameter growth; (3) simulation-driven data mixing using utility estimation; and (4) 4-bit quantization-aware training with self-distillation. We release our model weights and code to support future research in efficient on-device language models.

Arun Babu, Akshat Shrivastava, Armen Aghajanyan et al.

Semantic parsing using sequence-to-sequence models allows parsing of deeper representations compared to traditional word tagging based models. In spite of these advantages, widespread adoption of these models for real-time conversational use cases has been stymied by higher compute requirements and thus higher latency. In this work, we propose a non-autoregressive approach to predict semantic parse trees with an efficient seq2seq model architecture. By combining non-autoregressive prediction with convolutional neural networks, we achieve significant latency gains and parameter size reduction compared to traditional RNN models. Our novel architecture achieves up to an 81% reduction in latency on TOP dataset and retains competitive performance to non-pretrained models on three different semantic parsing datasets. Our code is available at https://github.com/facebookresearch/pytext

Sahithya Ravi, Patrick Huber, Akshat Shrivastava et al. · meta-ai

The emergence of Large Language Models (LLMs) has brought to light promising language generation capabilities, particularly in performing tasks like complex reasoning and creative writing. Consequently, distillation through imitation of teacher responses has emerged as a popular technique to transfer knowledge from LLMs to more accessible, Small Language Models (SLMs). While this works well for simpler tasks, there is a substantial performance gap on tasks requiring intricate language comprehension and creativity, such as humor generation. We hypothesize that this gap may stem from the fact that creative tasks might be hard to learn by imitation alone and explore whether an approach, involving supplementary guidance from the teacher, could yield higher performance. To address this, we study the effect of assigning a dual role to the LLM - as a "teacher" generating data, as well as a "critic" evaluating the student's performance. Our experiments on humor generation reveal that the incorporation of feedback significantly narrows the performance gap between SLMs and their larger counterparts compared to merely relying on imitation. As a result, our research highlights the potential of using feedback as an additional dimension to data when transferring complex language abilities via distillation.

Malak Mansour, Ahmed Aly, Bahey Tharwat et al.

Large Language Models (LLMs) such as GPT-4, trained on huge amount of datasets spanning multiple domains, exhibit significant reasoning, understanding, and planning capabilities across various tasks. This study presents the first-ever work in Arabic language integration within the Vision-and-Language Navigation (VLN) domain in robotics, an area that has been notably underexplored in existing research. We perform a comprehensive evaluation of state-of-the-art multi-lingual Small Language Models (SLMs), including GPT-4o mini, Llama 3 8B, and Phi-3 medium 14B, alongside the Arabic-centric LLM, Jais. Our approach utilizes the NavGPT framework, a pure LLM-based instruction-following navigation agent, to assess the impact of language on navigation reasoning through zero-shot sequential action prediction using the R2R dataset. Through comprehensive experiments, we demonstrate that our framework is capable of high-level planning for navigation tasks when provided with instructions in both English and Arabic. However, certain models struggled with reasoning and planning in the Arabic language due to inherent limitations in their capabilities, sub-optimal performance, and parsing issues. These findings highlight the importance of enhancing planning and reasoning capabilities in language models for effective navigation, emphasizing this as a key area for further development while also unlocking the potential of Arabic-language models for impactful real-world applications.

Ahmed Aly, Essam Mansour, Amr Youssef

Advanced Persistent Threats (APTs) are stealthy cyberattacks that often evade detection in system-level audit logs. Provenance graphs model these logs as connected entities and events, revealing relationships that are missed by linear log representations. Existing systems apply anomaly detection to these graphs but often suffer from high false positive rates and coarse-grained alerts. Their reliance on node attributes like file paths or IPs leads to spurious correlations, reducing detection robustness and reliability. To fully understand an attack's progression and impact, security analysts need systems that can generate accurate, human-like narratives of the entire attack. To address these challenges, we introduce OCR-APT, a system for APT detection and reconstruction of human-like attack stories. OCR-APT uses Graph Neural Networks (GNNs) for subgraph anomaly detection, learning behavior patterns around nodes rather than fragile attributes such as file paths or IPs. This approach leads to a more robust anomaly detection. It then iterates over detected subgraphs using Large Language Models (LLMs) to reconstruct multi-stage attack stories. Each stage is validated before proceeding, reducing hallucinations and ensuring an interpretable final report. Our evaluations on the DARPA TC3, OpTC, and NODLINK datasets show that OCR-APT outperforms state-of-the-art systems in both detection accuracy and alert interpretability. Moreover, OCR-APT reconstructs human-like reports that comprehensively capture the attack story.

Siddhant Arora, Haidar Khan, Kai Sun et al.

End-to-end speech-in speech-out dialogue systems are emerging as a powerful alternative to traditional ASR-LLM-TTS pipelines, generating more natural, expressive responses with significantly lower latency. However, these systems remain prone to hallucinations due to limited factual grounding. While text-based dialogue systems address this challenge by integrating tools such as web search and knowledge graph APIs, we introduce the first approach to extend tool use directly into speech-in speech-out systems. A key challenge is that tool integration substantially increases response latency, disrupting conversational flow. To mitigate this, we propose Streaming Retrieval-Augmented Generation (Streaming RAG), a novel framework that reduces user-perceived latency by predicting tool queries in parallel with user speech, even before the user finishes speaking. Specifically, we develop a post-training pipeline that teaches the model when to issue tool calls during ongoing speech and how to generate spoken summaries that fuse audio queries with retrieved text results, thereby improving both accuracy and responsiveness. To evaluate our approach, we construct AudioCRAG, a benchmark created by converting queries from the publicly available CRAG dataset into speech form. Experimental results demonstrate that our streaming RAG approach increases QA accuracy by up to 200% relative (from 11.1% to 34.2% absolute) and further enhances user experience by reducing tool use latency by 20%. Importantly, our streaming RAG approach is modality-agnostic and can be applied equally to typed input, paving the way for more agentic, real-time AI assistants.

Kai Zhang, Xinyuan Zhang, Ejaz Ahmed et al. · amazon-science

Accurate recall from large scale memories remains a core challenge for memory augmented AI assistants performing question answering (QA), especially in similarity dense scenarios where existing methods mainly rely on semantic distance to the query for retrieval. Inspired by how humans link information associatively, we propose AssoMem, a novel framework constructing an associative memory graph that anchors dialogue utterances to automatically extracted clues. This structure provides a rich organizational view of the conversational context and facilitates importance aware ranking. Further, AssoMem integrates multi-dimensional retrieval signals-relevance, importance, and temporal alignment using an adaptive mutual information (MI) driven fusion strategy. Extensive experiments across three benchmarks and a newly introduced dataset, MeetingQA, demonstrate that AssoMem consistently outperforms SOTA baselines, verifying its superiority in context-aware memory recall.

Darya Taratynova, Ahmed Aly, Numan Saeed et al.

Foundation models (FMs) are reshaping medical imaging, yet their application in echocardiography remains limited. While several echocardiography-specific FMs have recently been introduced, no standardized benchmark exists to evaluate them. Echocardiography poses unique challenges, including noisy acquisitions, high frame redundancy, and limited public datasets. Most existing solutions evaluate on private data, restricting comparability. To address this, we introduce CardioBench, a comprehensive benchmark for echocardiography FMs. CardioBench unifies eight publicly available datasets into a standardized suite spanning four regression and five classification tasks, covering functional, structural, diagnostic, and view recognition endpoints. We evaluate several leading FM, including cardiac-specific, biomedical, and general-purpose encoders, under consistent zero-shot, probing, and alignment protocols. Our results highlight complementary strengths across model families: temporal modeling is critical for functional regression, retrieval provides robustness under distribution shift, and domain-specific text encoders capture physiologically meaningful axes. General-purpose encoders transfer strongly and often close the gap with probing, but struggle with fine-grained distinctions like view classification and subtle pathology recognition. By releasing preprocessing, splits, and public evaluation pipelines, CardioBench establishes a reproducible reference point and offers actionable insights to guide the design of future echocardiography foundation models.

Hongda Jiang, Xinyuan Zhang, Siddhant Garg et al. · amazon-science

We introduce Memory-QA, a novel real-world task that involves answering recall questions about visual content from previously stored multimodal memories. This task poses unique challenges, including the creation of task-oriented memories, the effective utilization of temporal and location information within memories, and the ability to draw upon multiple memories to answer a recall question. To address these challenges, we propose a comprehensive pipeline, Pensieve, integrating memory-specific augmentation, time- and location-aware multi-signal retrieval, and multi-memory QA fine-tuning. We created a multimodal benchmark to illustrate various real challenges in this task, and show the superior performance of Pensieve over state-of-the-art solutions (up to 14% on QA accuracy).

Numan Saeed, Salma Hassan, Shahad Hardan et al.

We present a publicly available multimodal dataset for head and neck cancer research, comprising 1123 annotated Positron Emission Tomography/Computed Tomography (PET/CT) studies from patients with histologically confirmed disease, acquired from 10 international medical centers. All studies contain co-registered PET/CT scans with varying acquisition protocols, reflecting real-world clinical diversity from a long-term, multi-institution retrospective collection. Primary gross tumor volumes (GTVp) and involved lymph nodes (GTVn) were manually segmented by experienced radiation oncologists and radiologists following established guidelines. We provide anonymized NifTi files, expert-annotated segmentation masks, comprehensive clinical metadata, and radiotherapy dose distributions for a patient subset. The metadata include TNM staging, HPV status, demographics, long-term follow-up outcomes, survival times, censoring indicators, and treatment information. To demonstrate its utility, we benchmark three key clinical tasks: automated tumor segmentation, recurrence-free survival prediction, and HPV status classification, using state-of-the-art deep learning models like UNet, SegResNet, and multimodal prognostic frameworks.

Patrick Huber, Akshat Shrivastava, Ernie Chang et al. · meta-ai, mila

Sparse Mixture of Expert (MoE) models are popular foundational architectures at large scale, however, under-explored at smaller sizes. Here, we show how to enable Compact Sparse Mixture of Experts (CoSMoEs) for on-device inference. Specifically, we tackle the three main on-device dimensions: Quality, Memory and Latency. Along the quality axis, we show that in a fair evaluation (removing confounding factors) MoE architectures outperform FLOP-aligned dense models at on-device scale. We introduce weight-decomposed experts, further improving the MoE model performance. Regarding model memory and latency, we significantly improve model offloading efficiency and, in turn, reduce model inference latency.

Trang Le, Daniel Lazar, Suyoun Kim et al.

Spoken Language Understanding (SLU) is a critical component of voice assistants; it consists of converting speech to semantic parses for task execution. Previous works have explored end-to-end models to improve the quality and robustness of SLU models with Deliberation, however these models have remained autoregressive, resulting in higher latencies. In this work we introduce PRoDeliberation, a novel method leveraging a Connectionist Temporal Classification-based decoding strategy as well as a denoising objective to train robust non-autoregressive deliberation models. We show that PRoDeliberation achieves the latency reduction of parallel decoding (2-10x improvement over autoregressive models) while retaining the ability to correct Automatic Speech Recognition (ASR) mistranscriptions of autoregressive deliberation systems. We further show that the design of the denoising training allows PRoDeliberation to overcome the limitations of small ASR devices, and we provide analysis on the necessity of each component of the system.

Akshat Shrivastava, Shrey Desai, Anchit Gupta et al.

Task-oriented semantic parsing models have achieved strong results in recent years, but unfortunately do not strike an appealing balance between model size, runtime latency, and cross-domain generalizability. We tackle this problem by introducing scenario-based semantic parsing: a variant of the original task which first requires disambiguating an utterance's "scenario" (an intent-slot template with variable leaf spans) before generating its frame, complete with ontology and utterance tokens. This formulation enables us to isolate coarse-grained and fine-grained aspects of the task, each of which we solve with off-the-shelf neural modules, also optimizing for the axes outlined above. Concretely, we create a Retrieve-and-Fill (RAF) architecture comprised of (1) a retrieval module which ranks the best scenario given an utterance and (2) a filling module which imputes spans into the scenario to create the frame. Our model is modular, differentiable, interpretable, and allows us to garner extra supervision from scenarios. RAF achieves strong results in high-resource, low-resource, and multilingual settings, outperforming recent approaches by wide margins despite, using base pre-trained encoders, small sequence lengths, and parallel decoding.

Pooja Sethi, Denis Savenkov, Forough Arabshahi et al.

Improving the quality of Natural Language Understanding (NLU) models, and more specifically, task-oriented semantic parsing models, in production is a cumbersome task. In this work, we present a system called AutoNLU, which we designed to scale the NLU quality improvement process. It adds automation to three key steps: detection, attribution, and correction of model errors, i.e., bugs. We detected four times more failed tasks than with random sampling, finding that even a simple active learning sampling method on an uncalibrated model is surprisingly effective for this purpose. The AutoNLU tool empowered linguists to fix ten times more semantic parsing bugs than with prior manual processes, auto-correcting 65% of all identified bugs.

Shrey Desai, Akshat Shrivastava, Justin Rill et al.

Data efficiency, despite being an attractive characteristic, is often challenging to measure and optimize for in task-oriented semantic parsing; unlike exact match, it can require both model- and domain-specific setups, which have, historically, varied widely across experiments. In our work, as a step towards providing a unified solution to data-efficiency-related questions, we introduce a four-stage protocol which gives an approximate measure of how much in-domain, "target" data a parser requires to achieve a certain quality bar. Specifically, our protocol consists of (1) sampling target subsets of different cardinalities, (2) fine-tuning parsers on each subset, (3) obtaining a smooth curve relating target subset (%) vs. exact match (%), and (4) referencing the curve to mine ad-hoc (target subset, exact match) points. We apply our protocol in two real-world case studies -- model generalizability and intent complexity -- illustrating its flexibility and applicability to practitioners in task-oriented semantic parsing.

David Eriksson, Pierce I-Jen Chuang, Samuel Daulton et al.

When tuning the architecture and hyperparameters of large machine learning models for on-device deployment, it is desirable to understand the optimal trade-offs between on-device latency and model accuracy. In this work, we leverage recent methodological advances in Bayesian optimization over high-dimensional search spaces and multi-objective Bayesian optimization to efficiently explore these trade-offs for a production-scale on-device natural language understanding model at Facebook.

Shrey Desai, Ahmed Aly

Modern task-oriented semantic parsing approaches typically use seq2seq transformers to map textual utterances to semantic frames comprised of intents and slots. While these models are empirically strong, their specific strengths and weaknesses have largely remained unexplored. In this work, we study BART and XLM-R, two state-of-the-art parsers, across both monolingual and multilingual settings. Our experiments yield several key results: transformer-based parsers struggle not only with disambiguating intents/slots, but surprisingly also with producing syntactically-valid frames. Though pre-training imbues transformers with syntactic inductive biases, we find the ambiguity of copying utterance spans into frames often leads to tree invalidity, indicating span extraction is a major bottleneck for current parsers. However, as a silver lining, we show transformer-based parsers give sufficient indicators for whether a frame is likely to be correct or incorrect, making them easier to deploy in production settings.

Akshat Shrivastava, Pierce Chuang, Arun Babu et al.

An effective recipe for building seq2seq, non-autoregressive, task-oriented parsers to map utterances to semantic frames proceeds in three steps: encoding an utterance $x$, predicting a frame's length |y|, and decoding a |y|-sized frame with utterance and ontology tokens. Though empirically strong, these models are typically bottlenecked by length prediction, as even small inaccuracies change the syntactic and semantic characteristics of resulting frames. In our work, we propose span pointer networks, non-autoregressive parsers which shift the decoding task from text generation to span prediction; that is, when imputing utterance spans into frame slots, our model produces endpoints (e.g., [i, j]) as opposed to text (e.g., "6pm"). This natural quantization of the output space reduces the variability of gold frames, therefore improving length prediction and, ultimately, exact match. Furthermore, length prediction is now responsible for frame syntax and the decoder is responsible for frame semantics, resulting in a coarse-to-fine model. We evaluate our approach on several task-oriented semantic parsing datasets. Notably, we bridge the quality gap between non-autogressive and autoregressive parsers, achieving 87 EM on TOPv2 (Chen et al. 2020). Furthermore, due to our more consistent gold frames, we show strong improvements in model generalization in both cross-domain and cross-lingual transfer in low-resource settings. Finally, due to our diminished output vocabulary, we observe 70% reduction in latency and 83% reduction in memory at beam size 5 compared to prior non-autoregressive parsers.

Shrey Desai, Akshat Shrivastava, Alexander Zotov et al.

Task-oriented semantic parsing models typically have high resource requirements: to support new ontologies (i.e., intents and slots), practitioners crowdsource thousands of samples for supervised fine-tuning. Partly, this is due to the structure of de facto copy-generate parsers; these models treat ontology labels as discrete entities, relying on parallel data to extrinsically derive their meaning. In our work, we instead exploit what we intrinsically know about ontology labels; for example, the fact that SL:TIME_ZONE has the categorical type "slot" and language-based span "time zone". Using this motivation, we build our approach with offline and online stages. During preprocessing, for each ontology label, we extract its intrinsic properties into a component, and insert each component into an inventory as a cache of sorts. During training, we fine-tune a seq2seq, pre-trained transformer to map utterances and inventories to frames, parse trees comprised of utterance and ontology tokens. Our formulation encourages the model to consider ontology labels as a union of its intrinsic properties, therefore substantially bootstrapping learning in low-resource settings. Experiments show our model is highly sample efficient: using a low-resource benchmark derived from TOPv2, our inventory parser outperforms a copy-generate parser by +15 EM absolute (44% relative) when fine-tuning on 10 samples from an unseen domain.

Shrey Desai, Geoffrey Goh, Arun Babu et al.

The increasing computational and memory complexities of deep neural networks have made it difficult to deploy them on low-resource electronic devices (e.g., mobile phones, tablets, wearables). Practitioners have developed numerous model compression methods to address these concerns, but few have condensed input representations themselves. In this work, we propose a fast, accurate, and lightweight convolutional representation that can be swapped into any neural model and compressed significantly (up to 32x) with a negligible reduction in performance. In addition, we show gains over recurrent representations when considering resource-centric metrics (e.g., model file size, latency, memory usage) on a Samsung Galaxy S9.

Shrey Desai, Hongyuan Zhan, Ahmed Aly

The Lottery Ticket Hypothesis suggests large, over-parameterized neural networks consist of small, sparse subnetworks that can be trained in isolation to reach a similar (or better) test accuracy. However, the initialization and generalizability of the obtained sparse subnetworks have been recently called into question. Our work focuses on evaluating the initialization of sparse subnetworks under distributional shifts. Specifically, we investigate the extent to which a sparse subnetwork obtained in a source domain can be re-trained in isolation in a dissimilar, target domain. In addition, we examine the effects of different initialization strategies at transfer-time. Our experiments show that sparse subnetworks obtained through lottery ticket training do not simply overfit to particular domains, but rather reflect an inductive bias of deep neural networks that can be exploited in multiple domains.

Ahmed Aly, David Weikersdorfer, Claire Delaunay

This paper presents an evolutionary metaheuristic called Multiple Search Neuroevolution (MSN) to optimize deep neural networks. The algorithm attempts to search multiple promising regions in the search space simultaneously, maintaining sufficient distance between them. It is tested by training neural networks for two tasks, and compared with other optimization algorithms. The first task is to solve Global Optimization functions with challenging topographies. We found to MSN to outperform classic optimization algorithms such as Evolution Strategies, reducing the number of optimization steps performed by at least 2X. The second task is to train a convolutional neural network (CNN) on the popular MNIST dataset. Using 3.33% of the training set, MSN reaches a validation accuracy of 90%. Stochastic Gradient Descent (SGD) was able to match the same accuracy figure, while taking 7X less optimization steps. Despite lagging, the fact that the MSN metaheurisitc trains a 4.7M-parameter CNN suggests promise for future development. This is by far the largest network ever evolved using a pool of only 50 samples.

Ahmed Aly, Kushal Lakhotia, Shicong Zhao et al.

We introduce PyText - a deep learning based NLP modeling framework built on PyTorch. PyText addresses the often-conflicting requirements of enabling rapid experimentation and of serving models at scale. It achieves this by providing simple and extensible interfaces for model components, and by using PyTorch's capabilities of exporting models for inference via the optimized Caffe2 execution engine. We report our own experience of migrating experimentation and production workflows to PyText, which enabled us to iterate faster on novel modeling ideas and then seamlessly ship them at industrial scale.

Viral Thakar, Himani Saini, Walid Ahmed et al.

Asset monitoring in construction sites is an intricate, manually intensive task, that can highly benefit from automated solutions engineered using deep neural networks. We use Single-Shot Multibox Detector --- SSD, for its fine balance between speed and accuracy, to leverage ubiquitously available images and videos from the surveillance cameras on the construction sites and automate the monitoring tasks, hence enabling project managers to better track the performance and optimize the utilization of each resource. We propose to improve the performance of SSD by clustering the predicted boxes instead of a greedy approach like non-maximum suppression. We do so using Affinity Propagation Clustering --- APC to cluster the predicted boxes based on the similarity index computed using the spatial features as well as location of predicted boxes. In our attempts, we have been able to improve the mean average precision of SSD by 3.77% on custom dataset consist of images from construction sites and by 1.67% on PASCAL VOC Challenge.

Ahmed Aly, Joanne B. Dugan

Derivative-based optimization techniques such as Stochastic Gradient Descent has been wildly successful in training deep neural networks. However, it has constraints such as end-to-end network differentiability. As an alternative, we present the Accelerated Neuroevolution algorithm. The new algorithm is aimed towards physical robotic learning tasks following the Experiential Robot Learning method. We test our algorithm first on a simulated task of playing the game Flappy Bird, then on a physical NAO robot in a static Object Centering task. The agents successfully navigate the given tasks, in a relatively low number of generations. Based on our results, we propose to use the algorithm in more complex tasks.