Siva Reddy

Biographie

Siva Reddy est professeur adjoint en informatique et linguistique à l’Université McGill. Ses travaux portent sur les algorithmes qui permettent aux ordinateurs de comprendre et de traiter les langues humaines. Il a fait ses études postdoctorales avec le Stanford NLP Group. Son expertise inclut la construction de symboliques linguistiques et induites et de modèles d’apprentissage profond pour le langage.

Étudiants actuels

Vaibhav Adlakha

Doctorat - McGill

Parishad BehnamGhader

Maîtrise recherche - McGill

Doctorat - McGill

Collaborateur·rice de recherche - McGill

Verna Dankers

Postdoctorat - University of Edinburgh

Jiaqi Deng

Collaborateur·rice de recherche

Charbel El Feghali

Stagiaire de recherche - McGill

Desmond Elliott

Visiteur de recherche indépendant

Co-superviseur⋅e :

Yoshua Bengio

Jay Gala

Maîtrise recherche - McGill

Co-superviseur⋅e :

Collaborateur·rice de recherche

Hanseok Hanseok Oh

Collaborateur·rice alumni

Doctorat - McGill

Co-superviseur⋅e :

Timothy O'Donnell

Imene Kerboua

Collaborateur·rice de recherche - INSA Lyon, France

Doctorat - McGill

Superviseur⋅e principal⋅e :

Golnoosh Farnadi

Austin Kraft

Doctorat - McGill

Co-superviseur⋅e :

Doctorat - McGill

Zichao Li

Doctorat - McGill

Co-superviseur⋅e :

Jackie Cheung

Fengyuan Liu

Maîtrise recherche - McGill

Co-superviseur⋅e :

Dzmitry Bahdanau

Xing Han Lu

Doctorat - McGill

Maîtrise recherche - McGill

Doctorat - McGill

Postdoctorat - McGill

Marzia Nouri

Maîtrise recherche - McGill

Arkil Patel

Doctorat - McGill

Superviseur⋅e principal⋅e :

Collaborateur·rice de recherche - N/A

Ben Saine

Stagiaire de recherche - McGill

Dongchan Shin

Collaborateur·rice alumni

Karolina Ewa Stańczak

Collaborateur·rice alumni - McGill

Ivan Titov

Collaborateur·rice de recherche

Co-superviseur⋅e :

Yoshua Bengio

Ada Tur

Stagiaire de recherche - McGill

Doctorat - McGill

Collaborateur·rice alumni - McGill

Donghao Zeng

Stagiaire de recherche - McGill

Comment expliquer l’IA et s’assurer que cette explication est vraie? Les modèles mesurables de fidélité vous indiquent comment y parvenir

Billets de blogue

1 octobre 2024

par

Andrea Madsen

Siva Reddy

Sarath Chandar

Lire l'article

Publications

CUBE: A Standard for Unifying Agent Benchmarks

Alexandre Lacoste

Nicolas Gontier

Oleh Shliazhko

Aman Jaiswal

Kusha Sareen

Shailesh Nanisetty

Joan Cabezas

Manuel Del Verme

Omar G. Younis

Simone Baratta

Matteo Avalle

Imene Kerboua

Xing Han Lu

Elron Bandel

Michal Shmueli-Scheuer

Asaf Yehudai

Leshem Choshen

Jonathan Lebensold

Sean Hughes

Massimo Caccia … (voir 6 de plus)

Alexandre Drouin

Tao Yu

Yu Su

Graham Neubig

Dawn Song

The proliferation of agent benchmarks has created critical fragmentation that threatens research productivity. Each new benchmark requires s… (voir plus)ubstantial custom integration, creating an "integration tax" that limits comprehensive evaluation. We propose CUBE (Common Unified Benchmark Environments), a universal protocol standard built on MCP and Gym that allows benchmarks to be wrapped once and used everywhere. By separating task, benchmark, package, and registry concerns into distinct API layers, CUBE enables any compliant platform to access any compliant benchmark for evaluation, RL training, or data generation without custom integration. We call on the community to contribute to the development of this standard before platform-specific implementations deepen fragmentation as benchmark production accelerates through 2026.

2026-03-15

arXiv (prépublication)

arxiv.org

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)

arxiv.org

Hierarchical Retrieval at Scale: Bridging Transparency and Efficiency

Tianyi Chen

Valentina Zantedeschi

Information retrieval is a core component of many intelligent systems as it enables conditioning of outputs on new and large-scale datasets.… (voir plus) While effective, the standard practice of encoding data into high-dimensional representations for similarity search entails large memory and compute footprints, and also makes it hard to inspect the inner workings of the system. Hierarchical retrieval methods offer an interpretable alternative by organizing data at multiple granular levels, yet do not match the efficiency and performance of flat retrieval approaches. In this paper, we propose ReTreever, a tree-based method that makes hierarchical retrieval viable at scale by directly optimizing its structure for retrieval performance while naturally providing transparency through meaningful semantic groupings. Our method offers the flexibility to balance cost and utility by indexing data using representations from any tree level. We show that ReTreever delivers strong coarse (intermediate levels) and fine representations (terminal level), while achieving the highest retrieval accuracy at the lowest latency among hierarchical methods. These results demonstrate that this family of techniques is viable in practical applications.

2026-03-01

Trustworthy AI @ International Conference on Learning Representations (publié)

BRIDGE: Predicting Human Task Completion Time From Model Performance

Mila - Québec

AI Institute

McGill University

Polytechnique Montréal

Periodic Labs

Servicenow Research

Canada Cifar

AI Chair

Evaluating the real-world capabilities of AI systems requires grounding benchmark performance in human-interpretable measures of task diffic… (voir plus)ulty. Existing approaches that rely on direct human task completion time annotations are costly, noisy, and difficult to scale across benchmarks. In this work, we propose BRIDGE, a unified psychometric framework that learns the latent difficulty scale from model responses and anchors it to human task completion time. Using a two-parameter logistic Item Response Theory model, we jointly estimate latent task difficulty and model capability from model performance data across multiple benchmarks. We demonstrate that latent task difficulty varies linearly with the logarithm of human completion time, allowing human task completion time to be inferred for new benchmarks from model performance alone. Leveraging this alignment, we forecast frontier model capabilities in terms of human task length and independently reproduce METR's exponential scaling results, with the 50% solvable task horizon doubling approximately every 6 months.

2026-02-05

ArXiv (prépublication)

arxiv.org

The Markovian Thinker

Milad Aghajohari

Kamran Chitsaz

Amirhossein Kazemnejad

Reasoning LLMs suffer from quadratic compute growth as their context length increases, making reinforcement learning with verifiable rewards… (voir plus) (RLVR) and test-time scaling prohibitively expensive. Prior work has tried to lighten the computational burden by shortening reasoning traces through pruning, summarization, or multi-stage training, but these methods remain bound to quadratic costs. We introduce Delethink, a thinking algorithm that realizes the Markovian Thinking Paradigm. Instead of producing one long monolithic reasoning trace, Delethink thinks in a sequence of chunks, the Delethink trace. Each chunk continues reasoning by referring only to a fixed number of prior tokens, which functions as a Markovian state sufficient for progressing reasoning, while deleting the rest. This preserves continuity without carrying the quadratic baggage. As a result, compute scales linearly and peak memory remains constant. In experiments, we show that Delethink can be applied directly to off-the-shelf reasoning models ranging from

2025-12-31

International Conference on Learning Representations (Accept (Poster))

WebMMU: A Benchmark for Multimodal Multilingual Website Understanding and Code Generation

Christopher Pal

Juan A. Rodriguez

Sai Rajeswar

We present WebMMU, a multilingual benchmark that evaluates three core web tasks: (1) website visual question answering, (2) code editing inv… (voir plus)olving HTML/CSS/JavaScript, and (3) mockup-to-code generation. Unlike prior benchmarks that treat these tasks separately, WebMMU unifies them using expert-annotated, real-world web data to assess models'abilities in complex multi-step reasoning, precise element grounding, and functional UI comprehension and coding. Our evaluation shows that while multimodal large language models (MLLMs) perform well on basic information extraction, they struggle with reasoning and grounding, editing code to preserve functionality, and generating design-to-code that maintains hierarchy and supports multilingual content. These findings reveal key limitations in current MLLMs and underscore the need for improved multimodal and cross-lingual reasoning to build future web agents capable of automating diverse web development tasks.

2025-10-31

Conference on Empirical Methods in Natural Language Processing (publié)

SAFEARENA: Evaluating the Safety of Autonomous Web Agents

Ada Defne Tur

Esin Durmus

Karolina Stanczak

LLM-based agents are becoming increasingly proficient at solving web-based tasks. With this capability comes a greater risk of misuse for ma… (voir plus)licious purposes, such as posting misinformation in an online forum or selling illicit substances on a website. To evaluate these risks, we propose SafeArena, the first benchmark to focus on the deliberate misuse of web agents. SafeArena comprises 250 safe and 250 harmful tasks across four websites. We classify the harmful tasks into five harm categories -- misinformation, illegal activity, harassment, cybercrime, and social bias, designed to assess realistic misuses of web agents. We evaluate leading LLM-based web agents, including GPT-4o, Claude-3.5 Sonnet, Qwen-2-VL 72B, and Llama-3.2 90B, on our benchmark. To systematically assess their susceptibility to harmful tasks, we introduce the Agent Risk Assessment framework that categorizes agent behavior across four risk levels. We find agents are surprisingly compliant with malicious requests, with GPT-4o and Qwen-2 completing 34.7% and 27.3% of harmful requests, respectively. Our findings highlight the urgent need for safety alignment procedures for web agents. Our benchmark is available here: https://safearena.github.io

2025-10-05

Proceedings of the 42nd International Conference on Machine Learning (publié)

proceedings.mlr.press

How to Get Your LLM to Generate Challenging Problems for Evaluation

Arkil Patel

Dzmitry Bahdanau

The pace of evolution of Large Language Models (LLMs) necessitates new approaches for rigorous and comprehensive evaluation. Traditional hum… (voir plus)an annotation is increasingly impracticable due to the complexities and costs involved in generating high-quality, challenging problems, particularly for tasks such as long-context reasoning. Moreover, the rapid saturation of existing human-curated benchmarks by LLMs further necessitates the need to develop scalable and automatically renewable evaluation methodologies. In this work, we introduce **CHASE**, a unified framework to synthetically generate challenging problems using LLMs without human involvement. For a given task, our approach builds a hard problem in a bottom-up manner from simpler components. Moreover since we want to generate synthetic data for evaluation, our framework decomposes the generation process into independently verifiable sub-tasks, thereby ensuring a high level of quality and correctness. We implement CHASE to create evaluation benchmarks across three diverse domains: document-based question answering, repository-level code completion, and math reasoning. The performance of state-of-the-art LLMs on these synthetic benchmarks lies in the range of 40-60\% accuracy, thereby demonstrating the effectiveness of our framework at generating hard problems. Our experiments further reveal that the Gemini models significantly outperform other LLMs at long-context reasoning, and that the performance of all LLMs drastically drops by as much as 70\% when we scale up the context size to 50k tokens.

2025-09-22

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

Towards Democratizing LLMs: Investigating Multilingual Mixture-of-Experts Models

2025-09-21

NeurIPS.cc/2025/Workshop/WiML (publié)

The Promise of RL for Autoregressive Image Editing

Saba Ahmadi

Rabiul Awal

Ankur Sikarwar

Amirhossein Kazemnejad

Ge Ya Luo

Juan A. Rodriguez

Sai Rajeswar

Christopher Pal

Benno Krojer

Aishwarya Agrawal

While image generation techniques are now capable of producing high-quality images that respect prompts which span multiple sentences, the t… (voir plus)ask of text-guided image editing remains a challenge. Even edit requests that consist of only a few words often fail to be executed correctly. We explore three strategies to enhance performance on a wide range of image editing tasks: supervised fine-tuning (SFT), reinforcement learning (RL), and Chain-of-Thought (CoT) reasoning. In order to study all these components in one consistent framework, we adopt an autoregressive multimodal model that processes textual and visual tokens in a unified manner. We find RL combined with a large multi-modal LLM verifier to be the most effective of these strategies. As a result, we release EARL: Editing with Autoregression and RL, a strong RL-based image editing model that performs competitively on a diverse range of edits compared to strong baselines, despite using much less training data. Thus, EARL pushes the frontier of autoregressive multimodal models on image editing. We release our code, training data, and trained models at https://github.com/mair-lab/EARL.

2025-09-17

NeurIPS.cc/2025/Conference (poster)

Semantic change in adults is not primarily a generational phenomenon

Gaurav Kamath

Michelle Yang

Morgan Sonderegger

Dallas Card

A central question in the study of language change is whether or not such change is generational. If a language changes over time generation… (voir plus)-by-generation, the process looks as follows: New generations of speakers introduce innovations, while older speakers conserve their usage patterns, and the language changes as new generations replace older ones. At the opposite extreme, language change could be a zeitgeist phenomenon, in which changes are universally adopted by speakers simultaneously, regardless of age or generational cohort. This paper asks this question in the context of word meaning change. We analyze meaning change in over 100 words across more than 7.9 million U.S. congressional speeches, to observe whether, when a word sense rises or falls in prominence, adult speakers from different generations uniformly adopt it, or those from older generations conserve their prior usage. Using language model-based word sense induction methods, we identify different senses of each word, and then model the prevalence of each of these word senses as a function of time and speaker age. We find that most words show a small but statistically significant effect of speaker age; across almost 140 y of Congress, older speakers typically take longer than younger speakers to follow changes in word usage, but nevertheless do so within a few years. Our findings indicate that despite minor age-based differences, word meaning change among mature speakers is likely not a generational process, but rather a zeitgeist process, in which older adult speakers can readily adopt new word usage patterns.

2025-07-27

Proceedings of the National Academy of Sciences (publié)

AgentRewardBench: Evaluating Automatic Evaluations of Web Agent Trajectories

Xing Han Lu

Amirhossein Kazemnejad

Karolina Stanczak

Peter Shaw

Christopher J. Pal

Web agents enable users to perform tasks on web browsers through natural language interaction. Evaluating web agents trajectories is an impo… (voir plus)rtant problem, since it helps us determine whether the agent successfully completed the tasks. Rule-based methods are widely used for this purpose, but they are challenging to extend to new tasks and may not always recognize successful trajectories. We may achieve higher accuracy through human evaluation, but the process would be substantially slower and more expensive. Automatic evaluations with LLMs may avoid the challenges of designing new rules and manually annotating trajectories, enabling faster and cost-effective evaluation. However, it is unclear how effective they are at evaluating web agents. To this end, we propose AgentRewardBench, the first benchmark to assess the effectiveness of LLM judges for evaluating web agents. AgentRewardBench contains 1302 trajectories across 5 benchmarks and 4 LLMs. Each trajectory in AgentRewardBench is reviewed by an expert, who answers questions pertaining to the success, side effects, and repetitiveness of the agent. Using our benchmark, we evaluate 12 LLM judges and find that no single LLM excels across all benchmarks. We also find that the rule-based evaluation used by common benchmarks tends to underreport the success rate of web agents, highlighting a key weakness of rule-based evaluation and the need to develop more flexible automatic evaluations. We release the benchmark at: https://agent-reward-bench.github.io

2025-07-06

Conference on Language Modeling (accepté)