Nicolas Chapados

Christopher Pal

AI agents increasingly turn past experience into reusable artifacts such as code, workflows, and procedural memories. Reuse can improve effi… (voir plus)ciency, but it also creates a lifecycle reliability problem: artifacts that succeed once may fail under environment drift, underspecified tasks, or changing task distributions, especially in web automation. We introduce SKILL.nb, a framework for governing reusable agent workflows with evidence-calibrated lifecycle policies. SKILL.nb uses selective formalization: execution evidence decides which workflow steps should become executable code, which should remain natural-language guided, and when those choices should be revised. Workflows are stored as auditable, versioned notebooks that interleave natural-language guidance, multi-language executable cells, validation gates, fallback paths, and multimodal evidence such as outputs, screenshots, and error traces. At runtime, gate-conditioned execution lets each step run code when its gates validate, or fall back locally when drift invalidates the executable realization. On WebArena-Verified, SKILL.nb achieves 53.7% single-round success, improving over the strongest baseline by 3.9 percentage points. Across three re-executions, it retains 91.7% of initially successful tasks, 15.5 points above the next best method. Under bounded repair, it recovers 72.9% of subsequent failures while limiting post-repair regressions to 4.2%, compared with 15.0% to 17.0% for persistent baselines. It also leads on Mind2Web cross-website and cross-domain splits. In a GitLab migration test, SKILL.nb preserves performance when reusing frozen state learned on GitLab 15.7, with frozen-versus-fresh target-version gaps of -1.7 points on GitLab 16.11 and +0.6 points on GitLab 18.9. These results identify lifecycle governance and gate-conditioned execution as reliability axes beyond one-shot task success.

2026-06-05

arXiv (prépublication)

LLM2Vec-Gen: Generative Embeddings from Large Language Models

Parishad BehnamGhader

Fabian David Schmidt

LLM-based text embedders typically encode the semantic content of their input. However, embedding tasks require mapping diverse inputs to si… (voir plus)milar outputs. Typically, this input-output is addressed by training embedding models with paired data using contrastive learning. In this work, we propose a novel self-supervised approach, LLM2Vec-Gen, which adopts a different paradigm: rather than encoding the input, we learn to represent the model's potential response. Specifically, we add trainable special tokens to the LLM's vocabulary, append them to input, and optimize them to represent the LLM's response in a fixed-length sequence. Training is guided by the LLM's own completion for the query, along with an unsupervised embedding teacher that provides distillation targets. This formulation helps to bridge the input-output gap and transfers LLM capabilities such as safety alignment and reasoning to embedding tasks. Crucially, the LLM backbone remains frozen and training requires only unlabeled queries. LLM2Vec-Gen achieves state-of-the-art self-supervised performance on the Massive Text Embedding Benchmark (MTEB), improving by 9.3% over the best unsupervised embedding teacher. We also observe up to 43.2% reduction in harmful content retrieval and 29.3% improvement in reasoning capabilities for embedding tasks. Finally, the learned embeddings are interpretable and can be decoded into text to reveal their semantic content.

2026-03-10

arXiv (prépublication)

Grounding Computer Use Agents on Human Demonstrations

Aarash Feizi

Shravan Nayak

Xiangru Jian

Kevin Qinghong Lin

Kaixin Li

Rabiul Awal

Xing Han Lu

Johan Obando-Ceron

Juan A. Rodriguez

David Vázquez

Adriana Romero-Soriano

Reihaneh Rabbany

Perouz Taslakian

Christopher Pal

Spandana Gella

Sai Rajeswar

Building reliable computer-use agents requires grounding: accurately connecting natural language instructions to the correct on-screen eleme… (voir plus)nts. While large datasets exist for web and mobile interactions, high-quality resources for desktop environments are limited. To address this gap, we introduce GroundCUA, a large-scale desktop grounding dataset built from expert human demonstrations. It covers 87 applications across 12 categories and includes 56K screenshots, with every on-screen element carefully annotated for a total of over 3.56M human-verified annotations. From these demonstrations, we generate diverse instructions that capture a wide range of real-world tasks, providing high-quality data for model training. Using GroundCUA, we develop the GroundNext family of models that map instructions to their target UI elements. At both 3B and 7B scales, GroundNext achieves state-of-the-art results across five benchmarks using supervised fine-tuning, while requiring less than one-tenth the training data of prior work. Reinforcement learning post-training further improves performance. These results demonstrate the critical role of high-quality, expert-driven datasets in advancing general-purpose computer-use agents.

2025-12-31

International Conference on Learning Representations (Accept (Poster))

Grounding Computer Use Agents on Human Demonstrations

Aarash Feizi

Shravan Nayak

Xiangru Jian

Kevin Qinghong Lin

Kaixin Li

Rabiul Awal

Xing Han Lu

Johan Obando-Ceron

Juan A. Rodriguez

David Vázquez

Adriana Romero-Soriano

Reihaneh Rabbany

Perouz Taslakian

Christopher Pal

Spandana Gella

Sai Rajeswar

Building reliable computer-use agents requires grounding: accurately connecting natural language instructions to the correct on-screen eleme… (voir plus)nts. While large datasets exist for web and mobile interactions, high-quality resources for desktop environments are limited. To address this gap, we introduce GroundCUA, a large-scale desktop grounding dataset built from expert human demonstrations. It covers 87 applications across 12 categories and includes 56K screenshots, with every on-screen element carefully annotated for a total of over 3.56M human-verified annotations. From these demonstrations, we generate diverse instructions that capture a wide range of real-world tasks, providing high-quality data for model training. Using GroundCUA, we develop the GroundNext family of models that map instructions to their target UI elements. At both 3B and 7B scales, GroundNext achieves state-of-the-art results across five benchmarks using supervised fine-tuning, while requiring less than one-tenth the training data of prior work. Reinforcement learning post-training further improves performance, and when evaluated in an agentic setting on the OSWorld benchmark using o3 as planner, GroundNext attains comparable or superior results to models trained with substantially more data,. These results demonstrate the critical role of high-quality, expert-driven datasets in advancing general-purpose computer-use agents.

2025-11-09

arXiv (prépublication)

Malice in Agentland: Down the Rabbit Hole of Backdoors in the AI Supply Chain

Léo Boisvert

Abhay Puri

Chandra Kiran Reddy Evuru

Nazanin Sepahvand

Quentin Cappart

Alexandre Lacoste

Krishnamurthy (DJ) Dvijotham

Alexandre Drouin

The practice of fine-tuning AI agents on data from their own interactions--such as web browsing or tool use--, while being a strong general … (voir plus)recipe for improving agentic capabilities, also introduces a critical security vulnerability within the AI supply chain. In this work, we show that adversaries can easily poison the data collection pipeline to embed hard-to-detect backdoors that are triggerred by specific target phrases, such that when the agent encounters these triggers, it performs an unsafe or malicious action. We formalize and validate three realistic threat models targeting different layers of the supply chain: 1) direct poisoning of fine-tuning data, where an attacker controls a fraction of the training traces; 2) environmental poisoning, where malicious instructions are injected into webpages scraped or tools called while creating training data; and 3) supply chain poisoning, where a pre-backdoored base model is fine-tuned on clean data to improve its agentic capabilities. Our results are stark: by poisoning as few as 2% of the collected traces, an attacker can embed a backdoor causing an agent to leak confidential user information with over 80% success when a specific trigger is present. This vulnerability holds across all three threat models. Furthermore, we demonstrate that prominent safeguards, including two guardrail models and one weight-based defense, fail to detect or prevent the malicious behavior. These findings highlight an urgent threat to agentic AI development and underscore the critical need for rigorous security vetting of data collection processes and end-to-end model supply chains.

2025-10-02

arXiv (prépublication)

DRBench: A Realistic Benchmark for Enterprise Deep Research

Amirhossein Abaskohi

Tianyi Chen

Miguel Muñoz-Mármol

Curtis Fox

Amrutha Varshini Ramesh

Étienne Marcotte

Christopher Pal

Issam Hadj Laradji

We introduce DRBench, a benchmark for evaluating AI agents on complex, open-ended deep research tasks in enterprise settings. Unlike prior b… (voir plus)enchmarks that focus on simple questions or web-only queries, DRBench evaluates agents on multi-step queries (for example, ``What changes should we make to our product roadmap to ensure compliance with this standard?") that require identifying supporting facts from both the public web and private company knowledge base. Each task is grounded in realistic user personas and enterprise context, spanning a heterogeneous search space that includes productivity software, cloud file systems, emails, chat conversations, and the open web. Tasks are generated through a carefully designed synthesis pipeline with human-in-the-loop verification, and agents are evaluated on their ability to recall relevant insights, maintain factual accuracy, and produce coherent, well-structured reports. We release 15 deep research tasks across 10 domains, such as Sales, Cybersecurity, and Compliance. We demonstrate the effectiveness of DRBench by evaluating diverse DR agents across open- and closed-source models (such as GPT, Llama, and Qwen) and DR strategies, highlighting their strengths, weaknesses, and the critical path for advancing enterprise deep research. Code is available at https://github.com/ServiceNow/drbench.

2025-09-29

ArXiv (prépublication)

WebArena Verified: Reliable Evaluation for Web Agents

Amine El hattami

Megh Thakkar

Christopher Pal

Autonomous web agents increasingly operate in multi-step browser workflows, yet widely used benchmarks can misestimate performance due to un… (voir plus)derspecified goals and brittle checkers—challenges characteristic of normal benchmark maturation rather than flaws in the paradigm. We present WebArena Verified, a reproducible re-evaluation of WebArena that preserves its containerized environments while strengthening measurement. We audit all 812 tasks, repair misaligned evaluations and clarify ambiguous instructions; replace substring matching with type- and normalization-aware comparators; verify backend state for state-changing tasks; and adopt a structured JSON schema with explicit status codes for deterministic scoring. We provide improved results reporting with template-level macro averages, 95\% confidence intervals, and failure-mode breakdowns. We also introduce WebArena Verified Hard, a 137-task subset that retains difficult cases while reducing evaluation cost by 83\%. On the baseline agent we evaluated, it reduces false negatives by approximately 11\%. WebArena Verified remains drop-in compatible with minimal change to existing agents, supporting faithful and comparable progress. We release our code, data, and evaluation tools in our public repository.

2025-09-27

NeurIPS.cc/2025/Workshop/SEA (poster)

Beyond Naive Prompting: Strategies for Improved Zero-shot Context-aided Forecasting with LLMs

Arjun Ashok

Andrew Robert Williams

Vincent Zhihao Zheng

Irina Rish

Étienne Marcotte

Valentina Zantedeschi

Alexandre Drouin

Forecasting in real-world settings requires models to integrate not only historical data but also relevant contextual information, often ava… (voir plus)ilable in textual form. While recent work has shown that large language models (LLMs) can be effective context-aided forecasters via naïve direct prompting, their full potential remains underexplored. We address this gap with 4 strategies, providing new insights into the zero-shot capabilities of LLMs in this setting. ReDP improves interpretability by eliciting explicit reasoning traces, allowing us to assess the model's reasoning over the context independently from its forecast accuracy. CorDP leverages LLMs solely to refine existing forecasts with context, enhancing their applicability in real-world forecasting pipelines. IC-DP proposes embedding historical examples of context-aided forecasting tasks in the prompt, substantially improving accuracy even for the largest models. Finally, RouteDP optimizes resource efficiency by using LLMs to estimate task difficulty, and routing the most challenging tasks to larger models. Evaluated on different kinds of context-aided forecasting tasks from the CiK benchmark, our strategies demonstrate distinct benefits over naïve prompting across LLMs of different sizes and families. These results open the door to further simple yet effective improvements in LLM-based context-aided forecasting.

2025-09-22

BERT2S @ Neural Information Processing Systems (poster)

UI-Vision: A Desktop-centric GUI Benchmark for Visual Perception and Interaction

Shravan Nayak

Xiangru Jian

Kevin Qinghong Lin

Juan A. Rodriguez

Montek Kalsi

M. Tamer Özsu

Christopher Pal

Sai Rajeswar

Human Annotator

Autonomous agents that navigate Graphical User Interfaces (GUIs) to automate tasks like document editing and file management can greatly enh… (voir plus)ance computer workflows. While existing research focuses on online settings, desktop environments, critical for many professional and everyday tasks, remain underexplored due to data collection challenges and licensing issues. We introduce UI-Vision, the first comprehensive, license-permissive benchmark for offline, fine-grained evaluation of computer use agents in real-world desktop environments. Unlike online benchmarks, UI-Vision provides: (i) dense, high-quality annotations of human demonstrations, including bounding boxes, UI labels, and action trajectories (clicks, drags, and keyboard inputs) across 83 software applications, and (ii) three fine-to-coarse grained tasks—Element Grounding, Layout Grounding, and Action Prediction—with well-defined metrics to rigorously evaluate agents’ performance in desktop environments. Our evaluation reveals critical limitations in state-of-the-art models like UI-TARS-72B, including issues with understanding professional software, spatial reasoning, and complex actions like drag-and-drop. These findings highlight the challenges in developing fully autonomous computer-use agents. With UI-Vision, we aim to advance the development of more capable agents for real-world desktop tasks.

2025-07-14

International Conference on Machine Learning (Accept (poster))

proceedings.mlr.press

Spaced Scheduling for Large Language Model Training

Amine El hattami

Christopher Pal

2025-06-08

TMLR (accepté)

Silent Sabotage: Injecting Backdoors into AI Agents Through Fine-Tuning

Léo Boisvert

Abhay Puri

Chandra Kiran Reddy Evuru

Joshua Kazdan

Avinandan Bose

Quentin Cappart

Maryam Fazel

Sai Rajeswar

Jason Stanley

Alexandre Drouin

Krishnamurthy Dj Dvijotham

The rise of AI agents that can use tools, browse the web and interact with computers on behalf of a user, has sparked strong interest in imp… (voir plus)roving these capabilities by explicitly fine-tuning the LLMs/VLMs that power these agents. Several researchers have proposed collecting data by letting the agents interact with their environment (e.g., a computer operating system, the web or a collection of APIs exposed as tools), and improve agent performance by fine tuning on this data. In this work, we show that such data collection can be manipulated by adversaries to insert poisoned traces. By modifying just 5% of collected traces, adversaries can embed stealthy bad behaviors into agents—like leaking confidential user information whenever the tool or webpage exposes a trigger. Our results raise important security concerns in the development of AI agents, and underscore the importance of careful scrutiny of all data collection processes used to improve agentic AI.

2025-06-07

ICML.cc/2025/Workshop/WCUA (poster)

Context is Key: A Benchmark for Forecasting with Essential Textual Information

Arjun Ashok

Andrew R. Williams

Étienne Marcotte

Valentina Zantedeschi

Jithendaraa Subramanian

Alexandre Lacoste

Forecasting is a critical task in decision-making across numerous domains. While historical numerical data provide a start, they fail to con… (voir plus)vey the complete context for reliable and accurate predictions. Human forecasters frequently rely on additional information, such as background knowledge and constraints, which can efficiently be communicated through natural language. However, in spite of recent progress with LLM-based forecasters, their ability to effectively integrate this textual information remains an open question. To address this, we introduce "Context is Key" (CiK), a time-series forecasting benchmark that pairs numerical data with diverse types of carefully crafted textual context, requiring models to integrate both modalities; crucially, every task in CiK requires understanding textual context to be solved successfully. We evaluate a range of approaches, including statistical models, time series foundation models, and LLM-based forecasters, and propose a simple yet effective LLM prompting method that outperforms all other tested methods on our benchmark. Our experiments highlight the importance of incorporating contextual information, demonstrate surprising performance when using LLM-based forecasting models, and also reveal some of their critical shortcomings. This benchmark aims to advance multimodal forecasting by promoting models that are both accurate and accessible to decision-makers with varied technical expertise. The benchmark can be visualized at https://servicenow.github.io/context-is-key-forecasting/v0/.

2025-04-30

International Conference on Machine Learning (poster)